Does defense matter in OOTP6?

[darkhorse]

Originally posted by dougaiton
If pitchers left in an OF spot always hit .240 with 10 homers, or if batters left as a starter were in the region of a 5.00 ERA and with 100ks, this would be seen as a serious problem. To me, leaving anyone in a position should be a disaster.

How would I deal with it? Just like on-line leagues, use the defensive scale. Correct me if I'm wrong but it runs from SS-2B-CF all the way down to 1B-LF. Anyone trying to play down the scale can do using the normal system. Anyone not gets killed. It's not perfect, but it makes a lot more sense than presuming ML ability (even if very bad ML ability) at any position for any player.

Markus appears to be handicapped by growing up without baseball. As I understand it, he's only followed the game for seven years and didn't play it growing up. This makes some of the glaring faults in the game more understandable. Doesn't make me feel any better about them, however.

You can be so bad at a low-priority defensive position that it will hinder your career. Anyone remember the iron glove of first baseman Dick Stuart. Playing a first baseman at 2b or SS should have tragic consequences in OOTP, and it most certainly does NOT. Also, as many have noted, the AI has been dumbed down considerably regarding assigning position.

It really chaps the ass when you see things that worked in the past, go backwards in a new version of the game. Depth chart being another area of regress.

I'll say it as I always have. OOTP is a good game, the best on the market, but it needs to get better. Else, some bright fellow is going to make a better game, and these forums will become a ghost town. Much like the ITP forums.

[highandoutside]

Do we have any RL statistical evidence of what should be the impact of playing a player out of position? We should make sure that our assumptions about the drastic differences we think we should see are born out in reality....After all, didn't the people who do DIPS find that position players that were forced to pitch did not have radically different BABIP?

[dougaiton]

Well, that's kinda of the point. Because Frank Thomas would never, ever play an inning at short, OOTP has to make sure that its prohibitive for teams to do that. What we do know is that if there really was no disadvantage between Ruben Sierra and Enrique Wilson at short, we could be fairly sure that he would be starting (although that's maybe not the best franchise to pick). For example, we have no evidence as to how many ks, HRs or BBs (i.e. things the pitcher controls, unlike BABIP) a position player who has never pitched before would rack up in 35 starts, but its a fair assumption that it would be worse than the worst 10% of ML starters. What's happened here is that out-of-position players are slotted into the same scale that position players are judged by, which is judged on ML ability. Therefore, immediately an out-of-position player is gifted ML (although bad) ability at SS.

More to the point, I think research here has shown conclusively that after a few weeks/months, any player played at that position becomes a SS (even if a very bad one). To summarise the findings, defensive ability has a seemingly realistic affect on BABIP, in that bad players lead to a higher BABIP rating. However, the problem is that players who have never played the position are bumped on to the bottom of that scale. If OOTP is correct in saying that the worse the player, the higher the BABIP, is it wrong to assume that the difference between a bad defensive player and a player who has never played the position before would be significantly bigger than the difference between a players with a 20 range and 30 range?

[Big Train]

[QUOTE]Originally posted by darkhorse
[B]
It really chaps the ass when you see things that worked in the past, go backwards in a new version of the game. Depth chart being another area of regress.


Agreed, especially for those of us who play mainly in online leagues and have spent years building teams around principles that are suddenly no longer true.

>Would Frank Thomas or an athletic outfielder like Vladimir >Guerrero be able to play shortstop?

Thomas would be a disaster, I don't think you need any studies to prove that, any team who plays a guy like that at SS should be absolutely hammered.

The case of Vlad is more interesting, since he is athletic and has a good arm he *might* be able to learn to play at least a competent shortstop assuming he started learning it in the offseason. I think there's a good chance he wouldn't be able to be a very good shortstop or that the demands of a much more difficult defensive position which he is learning in the middle of his career would affect his offensive production. I think if most outfielders or firstbasemen could easily learn to play shortstop it would make sense for most teams to put them there since it's much easier to find a good offensive outfielder than a good offensive SS (even today). I think the fact that this hardly ever happens in real baseball shows that the transition is very difficult.

[darkhorse]

Originally posted by Big Train
I think there's a good chance he wouldn't be able to be a very good shortstop or that the demands of a much more difficult defensive position which he is learning in the middle of his career would affect his offensive production.

This is another area that the OOTP engine should model/incorporate. Playing a more difficult defensive position has a prohibitive effect on offense. Place a player at a position in which he is ill-suited and most likely his offense will dive and the risk of injury will skyrocket. Combine this with a more realistic impact on defense, and you would never see the AI or a human GM make this entirely unrealistic move. The price for this stupidity would be too high.

[obaslg]

I think we have separate questions:
1) How much should the difference between an average fielder and the best possible fielder affect the game?
2) Should the fielding ratings work in linear or geometric fashion - e.g., should a 90 be twice as good as a 45, or something else?
3) Given the answers to #1 and #2, how should the ratings of fictional players and roster sets be created?

1) This is very tough. Personally, I don't feel qualified to even offer an opinion, though presumably there's some research on this.
2) I strongly believe that the affect of ratings should be linear, or else the ratings won't be meaningful to us.
3) I think we have some flaws here. I think that the team has made a mistake in hanging on to the idea that a 50 should be average on a 100-point scale. As far as I can tell, this is where the whole ratings display problem came from. We also now know that the program, after importing a league that happens to contain very high ratings of one kind (say, many players with very high stuff), will gradually reduce the stuff ratings of all players to force the average to 50.

I think all of that is a mistake. A 100 should be considered the highest possible rating, presumably never before seen and never to be seen in the future - maybe equivalent to 70 predicted homers, for example. All other players would then be rated in relation to the 100. BTW, I say 70 “predicted” homers, because this would still allow for something amazing – like 70+ homers – to happen by random effect – i.e., a player with a rating for 60 predicted homers has a career year.

In the case of homers, the large majority of players would be well under a 50 rating – we would only predict 35+ homers for a small minority of players.

Fielding would be a little different. Consider SS. The large majority of PLAYERS would have a SS rating far less than half the theoretical maximum, but the large majority of current ML SS would have a rating far better than half.

This would leave plenty of room on the low end of the scale to put in a workable number for players playing out of position. Currently, a 5 rating isn't so bad if there are everyday ML SS with 20s, but if the average regular SS is a 75, a 5 is disastrous.

I realize that there's an intuitive appeal to the average being 50, but I don't think that it reflects reality. In reality, the average hitter doesn't come close to being half as good as the best hitter.

There’s a big interaction between #1 and #3 in my questions above. I think you would want to set the effect of defense by determining the effect of the best theoretical defender and worst theoretical defender at each position. You divide the difference in effect by 100 and you have the impact of each rating point. That would allow you to create roster sets from real stats.

To create fictional players, I would suggest that you find the worst regular player at each position in recent times, as measured by real stats, and determine his fielding rating from the system above. That would then be about the lowest rating of an fictional player created as a player at the position – e.g., if the worst real SS in the last 20 years computed to an OOTP rating of 40, then no fictional “SS” would be created with a rating less than 35.

However, another player moving to SS might have a rating far below 35 depending on his rating at his original position and which position it was. Alomar (or whoever’s considered a great defensive 2B right now – I’m completely out of touch) might become a 70 SS, Keith Hernandez a 40 (forgetting handedness), Willie Mays a 30, and Frank Thomas a 1.

[I guess I turned this into a bigger issue – I’d start a new thread for this, but everyone complains about that.]

[Skipaway]

Originally posted by darkhorse
Markus appears to be handicapped by growing up without baseball. As I understand it, he's only followed the game for seven years and didn't play it growing up. This makes some of the glaring faults in the game more understandable. Doesn't make me feel any better about them, however.

I don't think that's the reason. How easy is it to make mistakes in formulae and algorithms, especially in boundary conditions?

Markus made a similar mistake in pitcher rating in ITP. When a pitcher got 100 in everything, the pitcher would actually suck and give up tons of homeruns. Are you going to blame that on "growing up without baseball" too? Depth chart is also an algorithm issue, hardly got anything to do with baseball sense.

I blame both on not enough testing on boundary conditions.

[darkhorse]

Originally posted by Skipaway
Markus made a similar mistake in pitcher rating in ITP. When a pitcher got 100 in everything, the pitcher would actually suck and give up tons of homeruns. Are you going to blame that on "growing up without baseball" too?

Pfft. Like I would ever play that game!

[Henry]

Originally posted by Skipaway
I blame both on not enough testing on boundary conditions.

Not specifically directed at anyone, but this is an interesting comment considering what the board was like the weeks preceeding the game's release.

[darkhorse]

Originally posted by Henry
Not specifically directed at anyone, but this is an interesting comment considering what the board was like the weeks preceeding the game's release.

Hmm......wait and see how this board reacts if OOTP7 doesn't ship in 2005.

Slated for Fall 2005, IIRC.

Championship Manager invariably misses the initial ship date when a revamp of the engine occurs. The new EHM is two years and counting in the making. Would a similar delay with OOTP7 be unexpected?

I find it more interesting that OOTP6 didn't come out until after opening day. I'm sure it has nothing to do with Markus splitting his time between two projects.

[Skipaway]

Originally posted by Henry
Not specifically directed at anyone, but this is an interesting comment considering what the board was like the weeks preceeding the game's release.

Actually it might not be the issue of not enough testing time, but the testing philosophy. The first priority is always checking if the game feel realistic under normal/nominal settings.

Oddball settings are often ignored.

[darkhorse]

Originally posted by Skipaway
Actually it might not be the issue of not enough testing time, but the testing philosophy. The first priority is always checking if the game feel realistic under normal/nominal settings.

Beta tester?

[Skipaway]

Originally posted by darkhorse
Beta tester?

No, pure speculations from what Craig Scarborough said about "I can assure you that the fielding model is/was dead-on when it comes to the differences between a good fielder and a bad fielder" and the lack of difference between a bad fielder and someone who can't field.

And some of my experience with past OOTP5 and ITP bugs. Sometimes problems are VERY obvious if you try to set up unrealistic and unreasonable league/players.

[IatricSB]

Originally posted by Skipaway
I blame both on not enough testing on boundary conditions.

As far as depth chart goes, you'd be wrong coming to this conclusion.

[jbmagic]

maybe the best solution is for markus to make this game where if a player dont have a rating at a defensive position, it wont let him play there at all...it wont allowed you to move a player with no defensive rating at that position...


the player can only play a certain position only if he has rating. if he has no rating at a certain position , he has to learb it in spring training or teach him it in the minors....

[Big Train]

Originally posted by obaslg
I think we have separate questions:

Fielding would be a little different. Consider SS. The large majority of PLAYERS would have a SS rating far less than half the theoretical maximum, but the large majority of current ML SS would have a rating far better than half.

This would leave plenty of room on the low end of the scale to put in a workable number for players playing out of position. Currently, a 5 rating isn't so bad if there are everyday ML SS with 20s, but if the average regular SS is a 75, a 5 is disastrous.

I think you've hit on the really key point here. In the old system I always thought of A as great B as good C as average D as bad and E as so bad you'd be nuts to play him at a key defensive position.

I think if you took all major league shortstops and compared them the range would be relatively small. You might have the best guy in the 90-100 range and the worst guy in the 60-70 range.

The reason I think it has to be like this is because there is a *huge* difference between a relatively bad shortstop (who is none the less a trained shortstop) and someone playing out of position. I think you have to ask "What is the lowest level of defensive prowess an average manager would accept from a shortstop?" Maybe in the OOTP rating scale it would be 60, maybe 50 but it would NOT be 0 . If a shortsop was below the minimum acceptable standard he would be moved to another position quickly.

I think OOTP needs a system where there is a certain realistic range with the average being a lot higher than 50. Anyone playing someone below the minimum rating on this scale should be heavily penalized, anyone playing someone with a zero range should have a negative effect so great that they would finish last every year.

Sticking a power hitting outfielder at SS and having it have only a small negative effect on defense is ludicrous.

[Specs]

Originally posted by Big Train
Sticking a power hitting outfielder at SS and having it have only a small negative effect on defense is ludicrous.

I'm not suggesting that Mickey Stanley was an outstanding hitter, but you guys might want to look him up when thinking of placing an Ichiro, a Vlad, or a Johnny Damon at shortstop. If anyone is unfamiliar with Stanley, Tigers manager Mayo Smith played him (a natural outfielder and a damn good one) at short during the 1968 World Series to get his weak hitting defensive expert shortstops like Ray Oyler out of the lineup and all four of his talented outfielders in. The Tigers won the series.

[highandoutside]

fyi

Mickey Stanley's fielding stats

http://www.baseball-reference.com/s/stanlmi01.shtml

[darkhorse]

Originally posted by Specs
I'm not suggesting that Mickey Stanley was an outstanding hitter

If anyone is unfamiliar with Stanley, Tigers manager Mayo Smith played him (a natural outfielder and a damn good one) at short during the 1968 World Series to get his weak hitting defensive expert shortstops like Ray Oyler out of the lineup and all four of his talented outfielders in. The Tigers won the series.

He wasn't. Lifetime EQA of .251.

Stanley was a center-fielder. Moving a CF to middle-infield is not unheard of. As was pointed out, Stanley played 60-70 innings there during the regular season of '68.

[Specs]

He started nine games there all after the Tigers had clinched the pennant, so that Smith could see if that would work, and he played them rather well, with a range factor far above the "defensive specialists" on the team.