Clutch Hitting

Curtis · 09-19-2006, 11:20 AM

I am amazed that there is no clutch hitting rating in this game. As the boys in the booth say, it doesn't matter if you drive in a hundred, if you drive them in when you're up or down by five.Some players (A-Rod, Barry Bonds before his most recent playoff appearance) have a reputation for not delivering when it's nedded, even though they put up all star numbers season after season. Other players have clutch batting averages sixty points above their season averages (Jose Reyes of the current Mets), or else a reputation for delivering in the playoffs (Mr. September, Mr. October, Mr. November, Rivera).

Having said that we need this, I next need to say that we need a better clutch rating than Strat-o-Matic has. In Strat, anytime there's a man on second or third with two out, that's a clutch situation, regardless of what the score is. I realize that there is no really good measure of what constitutes a clutch situation, but 'man on second, two outs, you're up by five' isn't one of them, and 'score tied, man on second, no outs' is one. Late Inning Pressure Situation (LIPS) might be a good start, but it wouldn't take into account guys who suddenly turn it on during a pennant race or in the playoffs.

Of course, anybody can get hot for a couple of weeks, and the game does a fine job of depicting that, but some guys nearly always (seem to) rise to the occaision, year after year. In a way, you could accomodate this by giving players a 'seasonal tendency', which would be hidden and uneditable. (You'd have to figure it out by following the players' stats from year to year.) Some players deliver at close to the same pace all year round. Some flourish in the spring (ratings rise ten points on the 100 scale in spring training, April and May), some fade in the heat (ratings drop ten points in July and August), and some come alive in the fall.

Taking the spring guy as an example, he might play 'as if' he were fifteen points better in his ratings in March, ten better in April, five better in May, 'as printed' in June, and five points lower from the All Star Break through the playoffs, thus averaging him out to 'normal' for the full season. Mr. October might be three points low for the first five months, then from Labor Day through the Series he plays 'as if' ten points above. Mr. Dog Days plays five points better in April'June and September, but drops to ten below from the All Star Game through Labor Day.

Any of these things would apply equally well to pitchers. Right now it's hard to figure out why a given middle reliever converts into a very good or very bad closer, but a clutch rating could explain that. For pitchers you might want to add yet another rating, though — that for pitching with men on base. Some pitchers give up homers (or walks) at a normal rate, but mostly with the bases empty. Ex-Mets closer John Franco was semi-notorious for giving up solo homers when he had a two run lead, and sometimes the boys in the booth say the same about Martinez or a couple of the American League starters. Tom Seaver, Jim Palmer and Steve Carlton rarely gave up walks when first base was occupied.

This game is touted for its ability to keep track of obscure statistics, like how batters fare against specific pitchers, or how they hit with two strikes on them, but what's the point? Their ratings are the same regardless of the situation, so why keep situational stats? If their ratings changed because the game was on the line, or the temperature rose above 90 degrees, THEN it would make sense to keep track of that stuff.

KBLover · 09-19-2006, 01:16 PM

Sounds like the modifiers that were in BBPro. One for each month and one for C/L situations as well as ScPos situations.

I wouldn't mind this as long as it wasn't based on reputation. I mean, somehow A-Rod got the reputation for not delivering, but he has a career .935 OPS in the postseason.

This season he's got a .317 BA with RISP, with a .433 OBP and .531 SLG.

I don't think it should be uneditable. Suppose a roster creator wanted to use this? If it's just randomly generated whenever a player is, it could mess up how the roster would play, especially if the creator wanted to use a player's monthly splits for determining his monthly modifiers.

disposableheros · 09-19-2006, 04:17 PM

the first thing we need is a study that shows the actual measurable difference between non-clutch and clutch hitting stats.

that way Markus would be able to accurately code that change in ratings into the game.

HoustonReplay · 09-19-2006, 04:38 PM

I believe that any of the scientific (Sabermetric) studies done of clutch hitting have concluded that "IF it exists at all" "it is not consistent from year to year" and therefore that any apparent "ability" for clutch-hitting is not well understood, and does NOT stay with a hitter from year to year... therefore it may not be real.

the statistics for clutch-hitting can largely or totally be explained by mere randomness. (i.e., in my current replay of 1974, Ray Busse (a secondline player) has been quite CLUTCH ..... without any benefit from a game rating specific to clutch-hitting. Nonetheless Doug Rader remains the starting third-baseman.

IF OOTP were to include a significant clutch-hitting ability ... it should be made an OPTIONAL feature, otherwise the game is actually moving AWAY from being a simulation, by introducing a factor that most sabermetricians would consider to be non-existent. the same concept should apply to fantasy leagues .... having a clutch-hitting rating introduces in them a factor that while focused upon by broadcasters, is actually an illusion. If the game incorporates it, it should be as an Optional factor .. or else be a VERY minor influence on play by play events.

zuty · 09-23-2006, 12:08 PM

I think the game already represents clutch hitting just as much as real baseball does.

Look at any player in your league, and see how he does with men in scoring position in a close game. What? He hits .360 with an OPS of 1.125? He's a good clutch hitter! Oh, sorry -- he hits .218 with an OPS of .560? He's a bad clutch hitter.

Being a clutch hitter is based on whether or not you contribute in a situation that matters. I think that because some players in real life have done this and some players haven't does NOT mean there should be some added stat in the game.

Now, a feature that could be partially related to clutch hitting might be something like POISE. Maybe there's a poise rating, and that will influence a player's ability to survive on a big market team, or to win at the plate in the last at-bat of the World Series. Perhaps players with more poise would be less likely to argue a called third strike. Perhaps they would be more likely to win a bigger contract.

Rather than code the symptom (happening to go 0-5 in Game Seven of something), which is usually too small a cross section to be verifiable data, it would be better to code the core reason a player may or may not be able to do something.

Syd Thrift · 09-25-2006, 04:05 AM

It all depends on how you define "clutch". There definitely *is* definable ability that varies from player to player when it comes to hitting with men on base. It's not a huge effect, and it's not one held by the people you'd expect (i.e. Ichiro is "clutch" by this metric due to his slap-hitting that allows him to push hits through holes in the field, whereas Ryan Howard doesn't benefit all that much by guys being held on base), but it's certainly there.

Similarly, some guys hit better with 2 strikes than others. When you're doing a pitch-by-pitch kind of game, that can make a big difference: you can utilize a good bad-ball hitter with 1-run baseball tactics a whole lot better than you can a guy who turns into your pitcher when he gets 2 strikes on him. Is that "clutch"? Probably not in the way most people think of it, but it does exist and ought to be modeled.

tysok · 09-26-2006, 09:06 AM

Players have gotten a reputation for being clutch because they've delivered in abnormally huge moments, or failed in the same. A-Rod has gotten the reputation for not delivering because he's been in some extremely huge moments and struck out or grounded into a double play. Derek Jeter has a reputation for being a clutch player, Mr. November because he won the game with a home run. Reggie Jackson is Mr. October because he pounded 3 home runs in one world series game. David Ortiz is considered a clutch player because he's come up big in big situations.
Sabermetricians have looked at clutch stats and there's no reason to think it exists. The reason isn't because this player didn't come through in these situations and others did... it's because we (the people out here handing out these labels) remember the abnormally huge moments far better than the aveerage moments. To prove that it exists you have to take all the moments that could be clutch, however you want to define it, and look at how everyone compared.... we remember Derek Jeter hitting the home run in the World Series to become Mr. November, but that's an extremely small sample size. Anything to do with the world series would be, and especially that kind of situation where a home run would win it. So you go through the whole season and see that Derek Jeter failed as much as the next guy in "clutch" situations, or maybe he blew the roof off of it this year but the year before he went 0-25 in those situations.
That's not saying it doesn't exist, I believe it does to some degree. Some guys can perform when the pressure to do so is unbearable, and some crumple like an aluminum can. It's just that those situations are so small in number that there's not a large enough sample size in which to prove it, so you have to take all situations similar throughout the year... and decide that Derek Jeter didn't come up with that home run in the exact same situation in a game against Kansas City on April 4th... that wouldn't be anywhere near as pressure packed as the World Series, or near as important... but it fits in the definition of a clutch situation so his failure there has to figure with his huge success in the World Series.
The clutch ability can't be proved because of a sample size issue, and if you fix the sample size issue you prove it doesn't exist. Without being able to prove it's existence, it can't (at least shouldn't) be modeled in the game.

Joe Crede had a big hit last year in the postseason, Podsednik had a couple huge home runs last year in the postseason... these guys aren't labeled with the clutch tag, we just say they came through when they needed to.

sixto · 09-26-2006, 11:22 AM

Quote:

Originally Posted by Curtis Cook

Ex-Mets closer John Franco was semi-notorious for giving up solo homers when he had a two run lead

Just wondering what this has to do with clutch ability? Last time I checked a solo HR allowed with a 2-run lead does not result in a tie or a loss.

I think we need to approach this issue in terms of what we hope to get out of it. If a "normal" hitter gets a hit in clutch situations 3 out of 10 times, a "poor" clutch hitter gets a hit in clutch situations 2 out of 10 times, and a "great" clutch hitter 4 out of 10, in terms of gameplay, you won't really see that much of a change in the results - you will still see even the great player fail an average of 6 times out of 10.

Of course, randomization and small sample sizes can really skew the appearance of the results. If a clutch hitter goes 2-for-10 at some point, then he's statistically due to go 6-for-10, and say you watch one but sim the other? You'll be in the "this game suk it borken" crowd.

So to the people who want clutch in the game, I wonder, what specifically are you looking to see? Players who always get a hit with runners on 2nd and 3rd?

For me, the issue is more about the way the media covers clutch hitting than what actually happens. I think the game would be enhanced not by a changed game engine, but by better simulation of the media - the game noticing when players failed, or succeeded, in tough situations, and immortalizing that performance in game-generated articles.

Gastric ReFlux · 09-26-2006, 11:29 AM

Quote:

Originally Posted by sixto

For me, the issue is more about the way the media covers clutch hitting than what actually happens. I think the game would be enhanced not by a changed game engine, but by better simulation of the media - the game noticing when players failed, or succeeded, in tough situations, and immortalizing that performance in game-generated articles.

That would be an excellent development, if the news engine in the game generated stories in cases of batters delivering game-winning hits, and skewing more towards the times for teams near the top of the standings.

Curtis · 09-27-2006, 11:44 PM

KBLover: You want editable? Okay, we'll give you editable. I was thinking of this more for the historical crowd, and I wouldn't want someone to be able to edit Carl Yasterzemski into a Mr. October, but I suppose it would be all right (though subject to abuse) for fantasy players to be able to edit Joe Okamoto into a guy who hits .640 with runners in scoring position and two outs in the late innings of close games.

disposableheros: The first thing we need is a decent definition of what constitutes a clutch situation.

HoustonReplay: I agree that this should be optional, especially for fantasy rosters, but I would absolutely want this in any historical game I played in.

zuty: You're missing my point. As I said in my original post, "Of course, anybody can get hot for a couple of weeks, and the game does a fine job of depicting that, but some guys nearly always (seem to) rise to the occaision, year after year." In your examples I can't rely on these guys to be clutch because their performances have been a statistical aberration, not a reflection of true ability. As HoustonReplay said, even though Ray Busse has been clutch, Doug Rader gets the playing time, since he has a better 'card'.

Syd Thrift: Sorry to see you're dead, buddy — tough luck. I agree with your entire post.

tysok: Maybe I watched different playoffs than you did, but in my world Reggie Jackson came through in half a dozen post-seasons, not just one game. He became a LEGEND because of just one game, but he was Mr. October before the Yankees signed him. Likewise Babe Ruth and Yogi Berra (a legendary bad ball hitter, for the deceased Mr. Thrift) came through in the World Series year after year, even more than they did during the regular season, which is saying a lot (and which Lou Gehrig did not). I think the same could be said for Derek Jeter, and I HATE Derek Jeter. I didn't even remember him hitting a home run to win a series; what I remember is him ALWAYS getting the hit to keep a rally going during their four year streak, and doing amazing things in the field to prevent runs.

I think the rest of your post has a lot of merit. You and HoustonReplay brought up the lack of findings by Sabremetricians. Having studied statistics myself, I have to wonder what their criteria were for determining what is and what is not a clutch situation. As you pointed out, is a May game against a last place club as clutch a situation as an October game seven? I don't think a monthly modifier for historical players has much to justify it unless the player has four or five seasons under his belt, or his career has ended and you can look at it retrospectively. By the same token, a post-season modifier would be GIGO unless the player has twenty+ career post-season games under his belt, or (again) his career has ended.

sixto: You actually made my point, sort of. My point in mentioning John Franco is that this was a guy whose career home runs allowed per batter faced was probably fairly close to the major league average, especially for right-handed batters, but the homers he allowed never seemed to come when they made the difference in a game. Up until this season, Pedro Martinez had the same reputation. In a related vein, Keith Hernandez was talking in the booth a week or so ago and remembering reading in the paper when he was a kid about how Sandy Koufax would 'scatter seven, eight or a dozen hits and pitch a shutout or only allow one run. Is the ability to allow hits only when they won't hurt you a clutch ability? Enquiring minds want to know!

Gastric reFlux: Sorry, dude. You gave me nothing. Just kidding, but the improvement of the games wire reports is a good topic for another thread.

Everybody: Thank you for responding. I really didn't expect to get much feedback on this topic, and I'm glad y'all proved me wrong.

I think the biggest thing is trying to develop a definition for what constitutes 'clutch' in OOTP terms. For a batter that might be LIPS, but that wouldn't work well for modern day starters and long relievers, since they're so rarely around in the late innings.

The idea of a modifier for the playoffs grows out of the idea of seasonal or monthly modifiers, since EVERY at bat during the playoffs should be considered a clutch situation. The same might be said for a pennant race, but at what point should the game engine consider a pennant race to be in progress? Should you get a playoff modifier if you're up five games on Labor Day? If you're up five games with ten to play? Anytime you play the team just above or just below you in the standings? Or just during the playoffs? The playoff modifier would be completely separate from the clutch modifier, so both might apply at once.

disposableheros · 09-28-2006, 02:59 PM

Quote:

Originally Posted by Curtis Cook

disposableheros: The first thing we need is a decent definition of what constitutes a clutch situation.

and thats my point. if that definition doesnt exist in the real world, how is Markus going to code it into the game?

tysok · 09-29-2006, 09:41 AM

Quote:

Originally Posted by Curtis Cook

tysok: Maybe I watched different playoffs than you did, but in my world Reggie Jackson came through in half a dozen post-seasons, not just one game. He became a LEGEND because of just one game, but he was Mr. October before the Yankees signed him. Likewise Babe Ruth and Yogi Berra (a legendary bad ball hitter, for the deceased Mr. Thrift) came through in the World Series year after year, even more than they did during the regular season, which is saying a lot (and which Lou Gehrig did not). I think the same could be said for Derek Jeter, and I HATE Derek Jeter. I didn't even remember him hitting a home run to win a series; what I remember is him ALWAYS getting the hit to keep a rally going during their four year streak, and doing amazing things in the field to prevent runs.

I think the rest of your post has a lot of merit. You and HoustonReplay brought up the lack of findings by Sabremetricians. Having studied statistics myself, I have to wonder what their criteria were for determining what is and what is not a clutch situation. As you pointed out, is a May game against a last place club as clutch a situation as an October game seven? I don't think a monthly modifier for historical players has much to justify it unless the player has four or five seasons under his belt, or his career has ended and you can look at it retrospectively. By the same token, a post-season modifier would be GIGO unless the player has twenty+ career post-season games under his belt, or (again) his career has ended.

I'm sure I could be wrong, for most of Reggie Jackson's career I was too young to know what baseball was... let alone follow it.

But I believe he got the nickname in 1977, though it obviously fit him.
Derek Jeter won game 4 on November 1 2001 against the Diamondbacks. Obviously not a series winner, but a huge hit for the Yanks... one hit of very few that occurred, or may ever occur, in November... I still remember that call.

And actually, since Luis Gonzalez (I believe that's right) got the game winning hit to win the series... it seems the fancy nicknames like Mr. [Enter Month] are reserved for players on the Yankees.

Lou Gehrig appears to have done quite well in the World Series, that or he was 1920s or 30s version of Arod. He hit .361 and .731 SLG with 10 HRs and 35 RBIs in 7 series. While Berra put up realatively equal or lesser numbers in 14 series.
So Gehrig is either the early 20th century version of Arod, or it's selective rememberance... maybe Gehrig failed in one or two overly huge situations resulting in forgetting about the occasions he did come through..... I have no clue since I, apparently, had other pressing engagements at the time and didn't get to watch those World Series games in the 20s and 30s.

I'm not trying to be sarcastic, just going through what's known. The players you're mentioning as being clutch (Jackson, Jeter, Berra, Ruth) are all some of the greatest players to ever play the game. Are there any players that obviously have this clutch ability that aren't hall of famers, or future hall of famers? These types of guys, in the game, are such great players that they wouldn't need a clutch ability... if they get hot they put up outstanding numbers in the series... if they don't they put up average numbers.
Disposableheroes is correct, if experts can't prove that it exists, and how exactly it manifests itself, how is Markus supposed to put it in the game... keeping the game as realistic as possible?
If you only select what you would consider clutch situations in the postseason you may find guys that look to be excellent clutch hitters but really just got lucky. No player will ever reach a level where we normally consider their stats to be a solid representation of what they can do by taking only post season stats... it'd always fall in the "small sample size" category. That's why sabermetricians have to try and scope it out to include more situations... it's just not provable. Even without broadening the scope, you find that this year he may have done well and next year he doesn't.
Babe Ruth hit .118 in the 1922 WS... from just that base number he doesn't appear to have been very clutch. Podsednik and Crede had big numbers last year... if all 3 of those had been in the same year would Crede and Podsednik be considered clutch in the game while Ruth would be considered average? How would it be defined and executed?
The first thing that has to be done is to define it, then prove it with real data. Then it can be modeled... if it can't be proven (which to this point no ones been able to) then it can't be modeled.

Curtis · 09-29-2006, 10:30 AM

You keep insisting that one year makes a player clutch or not. I was saying that it takes repeated success over several years. Having now looked up most of the players we were refering to at Baseball Reference, I concede that the sabremetricians are probably correct. To whit, the following OBSs:

Babe Ruth — career — 1.164/41 WS games — 1.211 (and his ERA dropped from 2.28 career to 0.87 WS)
Lou Gehrig — career — 1.079/34 WS games — 1.208
Yogi Berra — career — .830/75 WS games — .811
Reggie Jackson — career — .846/27 WS games — 1.212
Derek Jeter — career — .847/115 post season games — .842
Barry Bonds — career — 1.053/48 post season games — .936

Conclusion? I was right about Ruth, Jackson and Bonds. I was wrong about Gehrig, Berra and Jeter. My memories WERE being clouded by sentiment, and any statistic that's only predictive half the time is pointless.

Does anyone have an idea whether the other 'well known' monthly trends have as little statistical grounding? And do you know where we could look it up to prove it?

(I'm still holding on to my LIPS, though. Hard as hell to make yourself understood without 'em.)

tysok · 09-29-2006, 12:59 PM

Quote:

Originally Posted by Curtis Cook

You keep insisting that one year makes a player clutch or not. I was saying that it takes repeated success over several years.

I understand that. The only problem is how many repeats would it take? Where is the line drawn between him being lucky and him being clutch? There's been occasions where a player seems to be god for 3 years then turns into crap... so you have to define a parameter, which is where the problems come in. It's easier to look at a body of work to decide... and if we decide he has to have 100 ABs in those situations it'd be likely that he'd end up with his "clutch" tag just in time for his decline and retirement...
It's obvious that you'd rather have Ruth, Bonds, or Jackson at the plate instead of Podsednik or Crede when it's get the big hit or go home time... but it's based more on the fact they've proved themselves to be far far better hitters than that they're "clutch".

I don't know about any other studies, but I think the monthly trends are more general than just month to month. I do think some players consistently start out slow and end strong, some end slowly but started strong. I always hear Greg Maddux starts out slow, Johann Santana puts up unbelievable numbers in the second half... but don't know of any studies.
I believe some pitchers get better as the game goes on, high 1st or 2nd ERA but very small later. Again, don't know any studies.
Both of those are based more on the information I hear from the broadcasters than on any stats I've looked up... they could be just as wrong as clutch performance that broadcasters enjoy talking about too.

Carlton_Willey · 09-29-2006, 01:10 PM

The biggest problem I have with 'clutch' ability is that both pitchers and batters are supposed to have it. So in measuring 'clutch' how do we know if a clutch pitcher is facing a clutch batter? Shouldn't that cancel out? In which case, how are we observers supposed to know that it was a confrontation between clutch players and not just ordinary guys?

And if clutch is there, shouldn't choking be to? And so how do we know if a pitcher striking out a batter with the bases loaded in the bottom on the ninth was the result of a clutch pitcher striking out an ordinary/not-as-clutch hitter, or an ordinary pitcher striking out a choking batter?

Personally, I don't believe in clutch but I think choking is true. I don't think anyone gets better under pressure, they simply choke less than the other guys on the field. The result looks like 'clutch' but its the opposite. A 99% non-choking Reggie Jackson thumped the 65% choking Bob Welch and other Dodger pitchers, is how I view the issue.

Curtis · 09-29-2006, 04:21 PM

"and if we decide he has to have 100 ABs in those situations it'd be likely that he'd end up with his "clutch" tag just in time for his decline and retirement..."

Absolutely true in real life, but in a historical simulation we'd already have the guy's career post-season numbers to judge by (or the roster makers would). And in a fantasy universe, the computer would know who was 'clutch' and who was 'choke' (thank you, Carlton) at the moment the character was created. The only people who wouldn't know would be the human players (a reason not to be able to edit it) and the scouts.

By the way, you won.

Quote:

Originally Posted by Carlton_Willey

The biggest problem I have with 'clutch' ability is that both pitchers and batters are supposed to have it. So in measuring 'clutch' how do we know if a clutch pitcher is facing a clutch batter? Shouldn't that cancel out? In which case, how are we observers supposed to know that it was a confrontation between clutch players and not just ordinary guys?

And if clutch is there, shouldn't choking be to? And so how do we know if a pitcher striking out a batter with the bases loaded in the bottom on the ninth was the result of a clutch pitcher striking out an ordinary/not-as-clutch hitter, or an ordinary pitcher striking out a choking batter?

By my definition, 'clutch' players do better than their career numbers in do-or-die situations, so in your example, Reggie Jackson would have to have a 143% non-choke rating. It's easier to just call it a 43% clutch rating.

You're right about the choking, though. I was always assuming it (with my Barry Bonds example), but wanted to express the concept more positively. (That's my story, and I'm sticking to it.)

As for what happens when a clutch pitcher meets a clutch hitter, all clutch ratings would do is raise specific ratings in do-or-die situations. In Jackson's case, his On Base was .101 higher in the World Series than during the season, so his contact, draw walk and possibly avoid strikeout ratings would rise. His Slugging Percentage rose .265, so his power (and likely gap) rating would soar. For the pitcher facing him, presumably his stuff and movement (and possibly control) would shift to an appropriate level.

I never thought that all players should have clutch (or choke) ratings. I assume that most players' performances are not markedly better or worse, adjusted for the level of competition, in the post-season than in the regular season. I thought that maybe 5-10% of players each would have either positive or negative clutch ratings, and everyone else would just keep their normal ratings.

Three other points: First, because the clutch/choke rating would be invisible and uneditable, the only way you'd know whether a clutch pitcher was facing a clutch hitter in any given matchup would be by your own personal observations of the players over time. (Or by a dramatic improvement in the play-by-play commentary.)

Second, when a clutch hitter faces a clutch pitcher, the pitcher wins. Just ask Carl Yatrezemski.

Third, I already lost this arguement, so please don't keep beating me while I'm down?

09-19-2006, 11:20 AM	#1
Curtis Hall Of Famer Join Date: Jul 2006 Location: Watertown, New York Posts: 4,567	Clutch Hitting I am amazed that there is no clutch hitting rating in this game. As the boys in the booth say, it doesn't matter if you drive in a hundred, if you drive them in when you're up or down by five.Some players (A-Rod, Barry Bonds before his most recent playoff appearance) have a reputation for not delivering when it's nedded, even though they put up all star numbers season after season. Other players have clutch batting averages sixty points above their season averages (Jose Reyes of the current Mets), or else a reputation for delivering in the playoffs (Mr. September, Mr. October, Mr. November, Rivera). Having said that we need this, I next need to say that we need a better clutch rating than Strat-o-Matic has. In Strat, anytime there's a man on second or third with two out, that's a clutch situation, regardless of what the score is. I realize that there is no really good measure of what constitutes a clutch situation, but 'man on second, two outs, you're up by five' isn't one of them, and 'score tied, man on second, no outs' is one. Late Inning Pressure Situation (LIPS) might be a good start, but it wouldn't take into account guys who suddenly turn it on during a pennant race or in the playoffs. Of course, anybody can get hot for a couple of weeks, and the game does a fine job of depicting that, but some guys nearly always (seem to) rise to the occaision, year after year. In a way, you could accomodate this by giving players a 'seasonal tendency', which would be hidden and uneditable. (You'd have to figure it out by following the players' stats from year to year.) Some players deliver at close to the same pace all year round. Some flourish in the spring (ratings rise ten points on the 100 scale in spring training, April and May), some fade in the heat (ratings drop ten points in July and August), and some come alive in the fall. Taking the spring guy as an example, he might play 'as if' he were fifteen points better in his ratings in March, ten better in April, five better in May, 'as printed' in June, and five points lower from the All Star Break through the playoffs, thus averaging him out to 'normal' for the full season. Mr. October might be three points low for the first five months, then from Labor Day through the Series he plays 'as if' ten points above. Mr. Dog Days plays five points better in April'June and September, but drops to ten below from the All Star Game through Labor Day. Any of these things would apply equally well to pitchers. Right now it's hard to figure out why a given middle reliever converts into a very good or very bad closer, but a clutch rating could explain that. For pitchers you might want to add yet another rating, though — that for pitching with men on base. Some pitchers give up homers (or walks) at a normal rate, but mostly with the bases empty. Ex-Mets closer John Franco was semi-notorious for giving up solo homers when he had a two run lead, and sometimes the boys in the booth say the same about Martinez or a couple of the American League starters. Tom Seaver, Jim Palmer and Steve Carlton rarely gave up walks when first base was occupied. This game is touted for its ability to keep track of obscure statistics, like how batters fare against specific pitchers, or how they hit with two strikes on them, but what's the point? Their ratings are the same regardless of the situation, so why keep situational stats? If their ratings changed because the game was on the line, or the temperature rose above 90 degrees, THEN it would make sense to keep track of that stuff.

09-19-2006, 04:17 PM	#3
disposableheros Hall Of Famer Join Date: Jun 2004 Posts: 4,332	the first thing we need is a study that shows the actual measurable difference between non-clutch and clutch hitting stats. that way Markus would be able to accurately code that change in ratings into the game. __________________ 2 Wild Cards, 11 Division Champs, 4 League Champs, 3 World Champs, and 3 Best GM awards Baseball Maelstrom - New York Mets - 180-149 .547 Corporate League Baseball - Coke Buzz - 889-649 .578 Western Hemisphere Baseball League - Santiago Saints - 672-793 .459 Record - 2428-2271 .517

09-19-2006, 01:16 PM	#2
KBLover All Star Reserve Join Date: Aug 2002 Posts: 580	Sounds like the modifiers that were in BBPro. One for each month and one for C/L situations as well as ScPos situations. I wouldn't mind this as long as it wasn't based on reputation. I mean, somehow A-Rod got the reputation for not delivering, but he has a career .935 OPS in the postseason. This season he's got a .317 BA with RISP, with a .433 OBP and .531 SLG. I don't think it should be uneditable. Suppose a roster creator wanted to use this? If it's just randomly generated whenever a player is, it could mess up how the roster would play, especially if the creator wanted to use a player's monthly splits for determining his monthly modifiers.

09-19-2006, 04:38 PM	#4
HoustonReplay Bat Boy Join Date: Jul 2006 Posts: 13	I believe that any of the scientific (Sabermetric) studies done of clutch hitting have concluded that "IF it exists at all" "it is not consistent from year to year" and therefore that any apparent "ability" for clutch-hitting is not well understood, and does NOT stay with a hitter from year to year... therefore it may not be real. the statistics for clutch-hitting can largely or totally be explained by mere randomness. (i.e., in my current replay of 1974, Ray Busse (a secondline player) has been quite CLUTCH ..... without any benefit from a game rating specific to clutch-hitting. Nonetheless Doug Rader remains the starting third-baseman. IF OOTP were to include a significant clutch-hitting ability ... it should be made an OPTIONAL feature, otherwise the game is actually moving AWAY from being a simulation, by introducing a factor that most sabermetricians would consider to be non-existent. the same concept should apply to fantasy leagues .... having a clutch-hitting rating introduces in them a factor that while focused upon by broadcasters, is actually an illusion. If the game incorporates it, it should be as an Optional factor .. or else be a VERY minor influence on play by play events.

09-23-2006, 12:08 PM	#5
zuty Minors (Double A) Join Date: Dec 2003 Posts: 173	I think the game already represents clutch hitting just as much as real baseball does. Look at any player in your league, and see how he does with men in scoring position in a close game. What? He hits .360 with an OPS of 1.125? He's a good clutch hitter! Oh, sorry -- he hits .218 with an OPS of .560? He's a bad clutch hitter. Being a clutch hitter is based on whether or not you contribute in a situation that matters. I think that because some players in real life have done this and some players haven't does NOT mean there should be some added stat in the game. Now, a feature that could be partially related to clutch hitting might be something like POISE. Maybe there's a poise rating, and that will influence a player's ability to survive on a big market team, or to win at the plate in the last at-bat of the World Series. Perhaps players with more poise would be less likely to argue a called third strike. Perhaps they would be more likely to win a bigger contract. Rather than code the symptom (happening to go 0-5 in Game Seven of something), which is usually too small a cross section to be verifiable data, it would be better to code the core reason a player may or may not be able to do something.

09-26-2006, 09:06 AM	#7
tysok Hall Of Famer Join Date: Aug 2003 Posts: 4,925	Players have gotten a reputation for being clutch because they've delivered in abnormally huge moments, or failed in the same. A-Rod has gotten the reputation for not delivering because he's been in some extremely huge moments and struck out or grounded into a double play. Derek Jeter has a reputation for being a clutch player, Mr. November because he won the game with a home run. Reggie Jackson is Mr. October because he pounded 3 home runs in one world series game. David Ortiz is considered a clutch player because he's come up big in big situations. Sabermetricians have looked at clutch stats and there's no reason to think it exists. The reason isn't because this player didn't come through in these situations and others did... it's because we (the people out here handing out these labels) remember the abnormally huge moments far better than the aveerage moments. To prove that it exists you have to take all the moments that could be clutch, however you want to define it, and look at how everyone compared.... we remember Derek Jeter hitting the home run in the World Series to become Mr. November, but that's an extremely small sample size. Anything to do with the world series would be, and especially that kind of situation where a home run would win it. So you go through the whole season and see that Derek Jeter failed as much as the next guy in "clutch" situations, or maybe he blew the roof off of it this year but the year before he went 0-25 in those situations. That's not saying it doesn't exist, I believe it does to some degree. Some guys can perform when the pressure to do so is unbearable, and some crumple like an aluminum can. It's just that those situations are so small in number that there's not a large enough sample size in which to prove it, so you have to take all situations similar throughout the year... and decide that Derek Jeter didn't come up with that home run in the exact same situation in a game against Kansas City on April 4th... that wouldn't be anywhere near as pressure packed as the World Series, or near as important... but it fits in the definition of a clutch situation so his failure there has to figure with his huge success in the World Series. The clutch ability can't be proved because of a sample size issue, and if you fix the sample size issue you prove it doesn't exist. Without being able to prove it's existence, it can't (at least shouldn't) be modeled in the game. Joe Crede had a big hit last year in the postseason, Podsednik had a couple huge home runs last year in the postseason... these guys aren't labeled with the clutch tag, we just say they came through when they needed to.

09-27-2006, 11:44 PM	#10
Curtis Hall Of Famer Join Date: Jul 2006 Location: Watertown, New York Posts: 4,567	KBLover: You want editable? Okay, we'll give you editable. I was thinking of this more for the historical crowd, and I wouldn't want someone to be able to edit Carl Yasterzemski into a Mr. October, but I suppose it would be all right (though subject to abuse) for fantasy players to be able to edit Joe Okamoto into a guy who hits .640 with runners in scoring position and two outs in the late innings of close games. disposableheros: The first thing we need is a decent definition of what constitutes a clutch situation. HoustonReplay: I agree that this should be optional, especially for fantasy rosters, but I would absolutely want this in any historical game I played in. zuty: You're missing my point. As I said in my original post, "Of course, anybody can get hot for a couple of weeks, and the game does a fine job of depicting that, but some guys nearly always (seem to) rise to the occaision, year after year." In your examples I can't rely on these guys to be clutch because their performances have been a statistical aberration, not a reflection of true ability. As HoustonReplay said, even though Ray Busse has been clutch, Doug Rader gets the playing time, since he has a better 'card'. Syd Thrift: Sorry to see you're dead, buddy — tough luck. I agree with your entire post. tysok: Maybe I watched different playoffs than you did, but in my world Reggie Jackson came through in half a dozen post-seasons, not just one game. He became a LEGEND because of just one game, but he was Mr. October before the Yankees signed him. Likewise Babe Ruth and Yogi Berra (a legendary bad ball hitter, for the deceased Mr. Thrift) came through in the World Series year after year, even more than they did during the regular season, which is saying a lot (and which Lou Gehrig did not). I think the same could be said for Derek Jeter, and I HATE Derek Jeter. I didn't even remember him hitting a home run to win a series; what I remember is him ALWAYS getting the hit to keep a rally going during their four year streak, and doing amazing things in the field to prevent runs. I think the rest of your post has a lot of merit. You and HoustonReplay brought up the lack of findings by Sabremetricians. Having studied statistics myself, I have to wonder what their criteria were for determining what is and what is not a clutch situation. As you pointed out, is a May game against a last place club as clutch a situation as an October game seven? I don't think a monthly modifier for historical players has much to justify it unless the player has four or five seasons under his belt, or his career has ended and you can look at it retrospectively. By the same token, a post-season modifier would be GIGO unless the player has twenty+ career post-season games under his belt, or (again) his career has ended. sixto: You actually made my point, sort of. My point in mentioning John Franco is that this was a guy whose career home runs allowed per batter faced was probably fairly close to the major league average, especially for right-handed batters, but the homers he allowed never seemed to come when they made the difference in a game. Up until this season, Pedro Martinez had the same reputation. In a related vein, Keith Hernandez was talking in the booth a week or so ago and remembering reading in the paper when he was a kid about how Sandy Koufax would 'scatter seven, eight or a dozen hits and pitch a shutout or only allow one run. Is the ability to allow hits only when they won't hurt you a clutch ability? Enquiring minds want to know! Gastric reFlux: Sorry, dude. You gave me nothing. Just kidding, but the improvement of the games wire reports is a good topic for another thread. Everybody: Thank you for responding. I really didn't expect to get much feedback on this topic, and I'm glad y'all proved me wrong. I think the biggest thing is trying to develop a definition for what constitutes 'clutch' in OOTP terms. For a batter that might be LIPS, but that wouldn't work well for modern day starters and long relievers, since they're so rarely around in the late innings. The idea of a modifier for the playoffs grows out of the idea of seasonal or monthly modifiers, since EVERY at bat during the playoffs should be considered a clutch situation. The same might be said for a pennant race, but at what point should the game engine consider a pennant race to be in progress? Should you get a playoff modifier if you're up five games on Labor Day? If you're up five games with ten to play? Anytime you play the team just above or just below you in the standings? Or just during the playoffs? The playoff modifier would be completely separate from the clutch modifier, so both might apply at once.

09-29-2006, 10:30 AM	#13
Curtis Hall Of Famer Join Date: Jul 2006 Location: Watertown, New York Posts: 4,567	You keep insisting that one year makes a player clutch or not. I was saying that it takes repeated success over several years. Having now looked up most of the players we were refering to at Baseball Reference, I concede that the sabremetricians are probably correct. To whit, the following OBSs: Babe Ruth — career — 1.164/41 WS games — 1.211 (and his ERA dropped from 2.28 career to 0.87 WS) Lou Gehrig — career — 1.079/34 WS games — 1.208 Yogi Berra — career — .830/75 WS games — .811 Reggie Jackson — career — .846/27 WS games — 1.212 Derek Jeter — career — .847/115 post season games — .842 Barry Bonds — career — 1.053/48 post season games — .936 Conclusion? I was right about Ruth, Jackson and Bonds. I was wrong about Gehrig, Berra and Jeter. My memories WERE being clouded by sentiment, and any statistic that's only predictive half the time is pointless. Does anyone have an idea whether the other 'well known' monthly trends have as little statistical grounding? And do you know where we could look it up to prove it? (I'm still holding on to my LIPS, though. Hard as hell to make yourself understood without 'em.)

09-29-2006, 01:10 PM	#15
Carlton_Willey Major Leagues Join Date: Apr 2005 Posts: 354	The biggest problem I have with 'clutch' ability is that both pitchers and batters are supposed to have it. So in measuring 'clutch' how do we know if a clutch pitcher is facing a clutch batter? Shouldn't that cancel out? In which case, how are we observers supposed to know that it was a confrontation between clutch players and not just ordinary guys? And if clutch is there, shouldn't choking be to? And so how do we know if a pitcher striking out a batter with the bases loaded in the bottom on the ninth was the result of a clutch pitcher striking out an ordinary/not-as-clutch hitter, or an ordinary pitcher striking out a choking batter? Personally, I don't believe in clutch but I think choking is true. I don't think anyone gets better under pressure, they simply choke less than the other guys on the field. The result looks like 'clutch' but its the opposite. A 99% non-choking Reggie Jackson thumped the 65% choking Bob Welch and other Dodger pitchers, is how I view the issue.