Sabermetrics are a bunch of malarkey . . .

[wlight1]

Originally Posted by Boom Boom

Sorry to dredge up an old post.

Well, it'd be awesome if you wanted to add something to the discussion. Rather than just.... bring it up again.

[mrbill]

There's nothing really to add. If I want to see what team was the best in a season, I look at wins. If I want to look at what hitter did the most individually for his team at the plate, I look at VORP. That tells me how much the team would have suffered if this guy was injured and they put a generic bench warmer in his place.

Anything besides that either disagrees completely with the premise of VORP or argues that VORP could be tweaked to weigh "highly leveraged" situations more to better estimate how much someone affected their team's chances of winning games.

[1998 Yankees]

Originally Posted by BigCity

Wow... can't tell you how many times this discussion has been on the board

Believe me, I will never bring it up again. But somebody else will.

[ctorg]

One thing that I think is an interesting thought regarding VORP is thinking about the concept of a "replacement player". Some teams have great benches featuring guys who could be starters on other teams. Let's say you have a team that has a 1B who puts up a VORP of 60 for the year, and they have a backup 1B who, if he were to have a full season, would put up a 40. Lucky team. So their VORP 60 guy then, I guess, is really only worth 20 to the team.

Whereas another team who also has a 60 VORP 1B but has a backup who would put up a 0 is really worth 60 to them.

To me this is an area that's important to look at when putting a team together. Not how much is a guy worth, but how much value he adds to your particular situation.

In the first team's case, if they already have the 40 VORP 1B and they look at the free agent market and see a 60 VORP 1B and a 40 VORP 3B, and they have a replacement level 3B, the 3B is worth more to them and they should go for the 3B over the 1B, all things being equal.

I have no idea if this post makes sense to anyone but me.

[jbergey22]

Makes a lot of sense Ctorg.

Your theory is basically if Albert Pujols has a VORP 60, while he backup has a VORP 40. Jimmy Rollins has a VORP 30 and his backup has a VORP 0. Despite having a 30 VORP less ranking Rollins would actually be more important. Correct?

As for the discussion, I like to hear both sides of argument. I like the 'metrics' as an accurate reflection of a player but I feel those in favor of the metrics overlook a few things. Juan Gonzalez was not the type of hitter that would put up great VORP numbers I am guessing but was a great cleanup hitter and drove in a lot of runs in his prime. He was up there swinging when guys were on base which is why he got RBI's. VORP gives too much credit for taking walks in situations where good hitters need to be driving in runs. I dont want Bonds up there trying to draw a walk with 2 outs and runners on 2nd and 3rd.

[ctorg]

The problem with my little system is that you have to kind of project the VORP of backup players. Since VORP isn't quite a rate stat, more opportunity means a higher VORP. So if a guy puts up a VORP of 10 in 100 AB, just as with any other stat, you don't know how well that projects. I mean, Gregg Jefferies's 1988 season was a pretty phenomenal 109 AB, but it didn't turn out to be very predictive.

So it's not really as simple as just looking at VORPs. There is always projection involved, which is where you can look at things like PECOTA and all that neat stuff.

[ctorg]

Originally Posted by jbergey22

Makes a lot of sense Ctorg.

Your theory is basically if Albert Pujols has a VORP 60, while he backup has a VORP 40. Jimmy Rollins has a VORP 30 and his backup has a VORP 0. Despite having a 30 VORP less ranking Rollins would actually be more important. Correct?

As for the discussion, I like to hear both sides of argument. I like the 'metrics' as an accurate reflection of a player but I feel those in favor of the metrics overlook a few things. Juan Gonzalez was not the type of hitter that would put up great VORP numbers I am guessing but was a great cleanup hitter and drove in a lot of runs in his prime. He was up there swinging when guys were on base which is why he got RBI's. VORP gives too much credit for taking walks in situations where good hitters need to be driving in runs. I dont want Bonds up there trying to draw a walk with 2 outs and runners on 2nd and 3rd.

One potential system to take situation into account (I don't know if this exists) would give lower value to a walk with 2 outs and runners on 2nd and 3rd than a leadoff walk. It would also try to factor in how often guys were up in various situations vs. how often those situations occur on average and adjust values accordingly. But there is still some discussion over how much a batter can really adjust to situations (e.g. "clutch"), so maybe the act of drawing a walk should be rated for value completely independent of the situation. It seems to me, though, that there may be some value in situational ratings.

[jbergey22]

Yeah this is a tricky situation. On one hand pitchers are likely a lot more afraid of Bonds in these 2 out situations therefore pitching around him which it turn hurts makes him less valuable in these situations because he is a better hitter where as Gonzalez would have likely seen somehing to swing at and more likely to drive in the runs.

Now you have to ask yourself despite it being obvious that Bonds is the better hitter. Who would you rather have up in this situation, the guy likely to get walked or the guy that will swing at whatever good ball he sees and is a known run producer? For argument sake lets assume the same hitter is on deck, Jeff Kent sounds good for some reason.

[RchW]

Originally Posted by ctorg

One thing that I think is an interesting thought regarding VORP is thinking about the concept of a "replacement player". Some teams have great benches featuring guys who could be starters on other teams. Let's say you have a team that has a 1B who puts up a VORP of 60 for the year, and they have a backup 1B who, if he were to have a full season, would put up a 40. Lucky team. So their VORP 60 guy then, I guess, is really only worth 20 to the team.

Whereas another team who also has a 60 VORP 1B but has a backup who would put up a 0 is really worth 60 to them.

To me this is an area that's important to look at when putting a team together. Not how much is a guy worth, but how much value he adds to your particular situation.

In the first team's case, if they already have the 40 VORP 1B and they look at the free agent market and see a 60 VORP 1B and a 40 VORP 3B, and they have a replacement level 3B, the 3B is worth more to them and they should go for the 3B over the 1B, all things being equal.

I have no idea if this post makes sense to anyone but me.

Maybe I'm missing the point but isn't the "replacement player" a virtual construction based on league wide averages for that season? It's not based on backup players that may be good or bad.

If so, then it doesn't matter that you have a good backup. VORP gives you an absolute ranking based on that seasons league averages. Your 60 VORP 1B is worth 60 no matter what his backup does.

However if you have good backup players, VORP will be just one of the stats you would use to create a better team. Either by finding another position for the good backup (DH for example), or trading for a comparable talent.

That's when I find VORP less useful. Many players have up and down careers so a GM must sometimes risk trading a currently good VORP player for someone who is lower but may have ratings/talent that should translate into good VORP. In my case I put more weight into contact and eye than power or K.

[ctorg]

Originally Posted by RchW

Maybe I'm missing the point but isn't the "replacement player" a virtual construction based on league wide averages for that season? It's not based on backup players that may be good or bad.

If so, then it doesn't matter that you have a good backup. VORP gives you an absolute ranking based on that seasons league averages. Your 60 VORP 1B is worth 60 no matter what his backup does.

However if you have good backup players, VORP will be just one of the stats you would use to create a better team. Either by finding another position for the good backup (DH for example), or trading for a comparable talent.

That's when I find VORP less useful. Many players have up and down careers so a GM must sometimes risk trading a currently good VORP player for someone who is lower but may have ratings/talent that should translate into good VORP. In my case I put more weight into contact and eye than power or K.

Yeah, that was kind of my point. VORP gives you how good a guy is relative to others at his position, but it doesn't necessarily tell you how much value he added to his team, because he's compared against an artificial construction: a replacement player. The player who actually replaces him (the backup) may be significantly better than a replacement player, which actually brings the amount of value that the starter adds to the team down.

Using OOTP ratings, let's say you've got a guy who has an overall value of 75 at 1B and a guy with an overall value of 45 at SS. Among the free agents is a 2B with a value of 79 and a SS with a value of 60. All else being equal, the SS with the 60 is actually worth more to your team because he will add more value above the backup.

That's all I'm saying. It's no big aha! or anything. It's why when you have two good players at one position you trade one away. You can add more value by filling holes than by upgrading at positions where you are already good.

[Left-handed Badger]

Originally Posted by ctorg

One thing that I think is an interesting thought regarding VORP is thinking about the concept of a "replacement player". Some teams have great benches featuring guys who could be starters on other teams. Let's say you have a team that has a 1B who puts up a VORP of 60 for the year, and they have a backup 1B who, if he were to have a full season, would put up a 40. Lucky team. So their VORP 60 guy then, I guess, is really only worth 20 to the team.

Whereas another team who also has a 60 VORP 1B but has a backup who would put up a 0 is really worth 60 to them.

To me this is an area that's important to look at when putting a team together. Not how much is a guy worth, but how much value he adds to your particular situation.

In the first team's case, if they already have the 40 VORP 1B and they look at the free agent market and see a 60 VORP 1B and a 40 VORP 3B, and they have a replacement level 3B, the 3B is worth more to them and they should go for the 3B over the 1B, all things being equal.

I have no idea if this post makes sense to anyone but me.

I thought of this last night, when this thread re-awakened.


But, this does make a case that VORP is just as subjective as RBI. Though, I am not getting into that again. I'll just let this be the "VORP appreciation" thread from here on out.

[Left-handed Badger]

Originally Posted by ctorg

The problem with my little system is that you have to kind of project the VORP of backup players. Since VORP isn't quite a rate stat, more opportunity means a higher VORP. So if a guy puts up a VORP of 10 in 100 AB, just as with any other stat, you don't know how well that projects. I mean, Gregg Jefferies's 1988 season was a pretty phenomenal 109 AB, but it didn't turn out to be very predictive.

So it's not really as simple as just looking at VORPs. There is always projection involved, which is where you can look at things like PECOTA and all that neat stuff.

Please, dont add no more stats to this. You guys have already overwhelmed the minds of us old-schoolers.

[wlight1]

Originally Posted by ctorg

Yeah, that was kind of my point. VORP gives you how good a guy is relative to others at his position, but it doesn't necessarily tell you how much value he added to his team, because he's compared against an artificial construction: a replacement player. The player who actually replaces him (the backup) may be significantly better than a replacement player, which actually brings the amount of value that the starter adds to the team down.

This is why David Ortiz's supporters had no legit argument for his MVP candidacy last season, in my opinion. If Alex Rodriguez got run over by a bus, the Yankees would have had Miguel Cairo or somebody out there. If Ortiz got smashed by an asteroid, the Sox would have been able to start a fairly good hitter in his place.

But yes, VORP's 'RP' isn't intended as a literal "replacement player". Again, VORP is intended to measure a player's skill -- in a vacuum -- at helping your team win baseball games. It doesn't (and shouldn't) impact Jim Thome's inherent value as a baseball player in '05-'06 that the Phillies have Ryan Howard waiting in the wings. It does (and, obviously, should) impact his real-world value to their team -- which is why they traded him.

But the fact that (numbers pulled out of arse for argument's sake) Thome had a (say) projected 55-VORP value that season and Howard had (say) projected at 50 doesn't mean Thome's Value Over (Actual) Replacement is 5. That's why VORP is a stat in a vacuum. He's a 55-VORP player, which is why the White Sox were willing to give up Rowand and Gio Gonzalez for him at the time (although now that I've come this far, bad example, because by acquiring Thome they were only upgrading at the DH position, likely not worth the downgrade from Roward to Brian Anderson, but that's neither here nor there. Anyway if they hadn't inexplicably resigned Konerko it'd have been a very different story).

Weighing your players and possible acquisitions against real-world alternatives is the job of an MLB GM. VORP's is to help demonstrate (by measuring against an arbitrary, imaginary constant) which of them, when placed on your team, would theoretically help you win more ballgames.

[Bonedwarf]

Every time I see this thread the title makes me chuckle.

[1998 Yankees]

Originally Posted by Bonedwarf

Every time I see this thread the title makes me chuckle.

Happy to oblige. And my spelling was correct, despite some naysayer opinions.

[ctorg]

Originally Posted by wlight1

But the fact that (numbers pulled out of arse for argument's sake) Thome had a (say) projected 55-VORP value that season and Howard had (say) projected at 50 doesn't mean Thome's Value Over (Actual) Replacement is 5. That's why VORP is a stat in a vacuum. He's a 55-VORP player, which is why the White Sox were willing to give up Rowand and Gio Gonzalez for him at the time (although now that I've come this far, bad example, because by acquiring Thome they were only upgrading at the DH position, likely not worth the downgrade from Roward to Brian Anderson, but that's neither here nor there. Anyway if they hadn't inexplicably resigned Konerko it'd have been a very different story).

Yes, VORP is useful because it basically gives you a guy's value relative to others at his position around the league. It is great for that.

When it comes to trades, a team should only trade if it is going to create an overall upgrade in itself. This seems obvious, but the nature of an upgrade isn't always straightforward. If you already have a pretty good 1B and you trade your not-too-bad CF for a superstar one, leaving you with a replacement-level player in CF, you'd better make sure the upgrade at 1B is more than the downgrade at CF.

Naturally there are other considerations, like age and contract, but going on the assumption that a team is in a sort of "win now" mode where they will spend what they need to in order to win, any trade made should produce a net improvement.

Like I said, this seems obvious, but sometimes you see teams make deals to get the big name guy that they don't need as much as the not-bad guy they gave up for him.

[spark240]

Wow. Several months away from OOTP forums, and this is the first thread I come upon.

Originally Posted by Jestre

Is a hit or walk that does not lead to a run of any value????VORP and RC think so, they give value to hits and walks that do not effect real runs.

Is a run scored or run batted in that does not lead to a win of any value? R and RBI think so, they give value to runs scored and runs batted in that do not effect real wins.

Is a win that does not lead to a World Series ring of any value? W thinks so, it gives value to wins that do not effect real World Series rings.

Obviously, the only statistic that matters is WSR.

[voxpoptart]

Originally Posted by spark240

Is a run scored or run batted in that does not lead to a win of any value? R and RBI think so, they give value to runs scored and runs batted in that do not effect real wins.

Is a win that does not lead to a World Series ring of any value? W thinks so, it gives value to wins that do not effect real World Series rings.

Obviously, the only statistic that matters is WSR.

I think you're trying to be funny, and in a just world, you would succeed in being funny. But half the professional sports columnists in the country are reading your words as straight gospel.

*****
To go back to the original topic: real life Runs Created formulas do currently include adjustments for clutch performance ("clutch" situations being runners-on-base, close-and-late, and leading-off-the-inning, which is a far more important clutch situation than people recognize without studying the issue). I don't know how the RC formulas adjust, because I don't care enough, but I know that they do. I assume Heinsohn doesn't program the game to remember situational hitting in enough detail to inform the RC calculations, that's all.

[injury log]

Originally Posted by 1998 Yankees

Happy to oblige. And my spelling was correct, despite some naysayer opinions.

Almost correct- you spelled 'is' wrong (you spelled it 'are'). 'Sabermetrics' is a singular mass noun.

[gglennwill]

OK guys, let's settle this argument once and for all. I want you to take the guys with the highest RBIs in your last league at each position, and create a team. Then take the guys with the highest VORP at each position, and create another team. Now play 1000 games, one team against the other. There you have it! Argument settled!