You're the GM (Game)

[czman]

Keep in mind that he took less to play for the Cubs. He clearly wants to be in Chicago. So unless things go terribly wrong, let's not assume he's eager to opt out of a deal with a team that he clearly wants to play for.

It is a 3 year deal and the Cubs paid him the most in those 3 years. The only way it is not a 3 year deal is he falls off the face the the earth.

 

[brett05]

Well in that case I showed you the numbers. For the entire 2015 season, every team, there was a 1.8 game variance on WAR to actual wins of .075%. The highest variance on an individual team WAR was 10% on your Angels example. If the number in the largest sample size, the totality of MLB, has a .075% variance and in the smallest individual components the largest variance is 10% why you can't accede to the point? In every meaningful measure I've proven the correlation to a minimum of 90%. You keep saying to show facts and I keep showing them to you without success. Do you not believe 90% to be a statistically significant correlation?

You took one team and then one year. And it looks good. Now expand that to the history of WAR. It will not be the same. Again, it is showing post, not projection.

 

[brett05]

Brett doesn't believe in odds or projections, it's viewed as witchcraft or something probably.

You're clearly clueless on how sports gambling works.

 

[brett05]

Vegas has the Cubs winning the NL at 7:2. 6:1 tied with SFG for the series.

Top AL team is BoSox.

My belief is the Dodgers got weaker which opens up the west for SFG to win it. That team is experienced in the play offs. That gives them the edge in the odds of winning if they get there. Almost stupid to bet against them in a series.

But the fact the Cubs are even odds to the play offs just speaks to their talent level. Seeing how they are the odds top winning team in the NL really says the odds makers are seeing this team being a regular season juggernaut but their lack of winning it all made them fall back to the Giants who just win WS's.

No it doesn't.

 

[brett05]

Kinda right. Cubs offered more per year with a opt out. Add to it it was the young core that pushed him. He felt better in that situation then being in turn over haul over his contract.

Sure the whole winning the first time stuff is there but you can't expect these guys to be fan boys of the team and all

Kinda right as well. He has two opt outs. He not only got more per year but great control over his future as well as the ability to land another nice contract by getting out two years earlier then the ten year contracts he turned down.

 

[TC in Mississippi]

You took one team and then one year. And it looks good. Now expand that to the history of WAR. It will not be the same. Again, it is showing post, not projection.

Yet you continue to argue with me that you're not arguing projection. At some point I'm going to go back 10 years with the same exercise and see the results. If the minimum correlation is still 90% or even 85% or better are you still going to argue with me? The fact is that you're talking about a tool. There is no magical projection out there. Not WAR, not Steamer, not Bill James, not ZiPS not anything and yet we love all of those things as baseball fans because they're interesting. They hone arguments. Statistical correllations and probability of outcomes are fascinating, they aren't profound nor do they pretend to be.

 

[brett05]

Yet you continue to argue with me that you're not arguing projection. At some point I'm going to go back 10 years with the same exercise and see the results. If the minimum correlation is still 90% or even 85% or better are you still going to argue with me? The fact is that you're talking about a tool. There is no magical projection out there. Not WAR, not Steamer, not Bill James, not ZiPS not anything and yet we love all of those things as baseball fans because they're interesting. They hone arguments. Statistical correllations and probability of outcomes are fascinating, they aren't profound nor do they pretend to be.

And I am ok with projections. When you made it dogma I have issues. That's the point. You want to do projections, try the projections for the 2015 season and see how accurate it was to what really happened.

 

[DanTown]

And I am ok with projections. When you made it dogma I have issues. That's the point. You want to do projections, try the projections for the 2015 season and see how accurate it was to what really happened.

That's not how to measure projections. Due to the fact that teams make trades, certain players get hurt, etc. Also, you'd have to look at a significantly larger sample than one year to argue that predictions (And whatever method) are ineffective.

 

[TC in Mississippi]

And I am ok with projections. When you made it dogma I have issues. That's the point. You want to do projections, try the projections for the 2015 season and see how accurate it was to what really happened.

I never presented it as dogma. Here was my original post that started the conversation:

Carson Cistulli of Fangraphs has posted a Cubs depth chart with fWAR values. It's kind of insane and if my math is right the projection is 97 wins. https://twitter.com/cistulli/status/681879000931786752

After that I mentioned that I didn't realize the correlation between WAR and actual wins was as high as it was until someone pointed it out to me with the Cubs example. Since then we've been mired in semantics and a back and forth on whether we were talking about projections or correlations to actual data. At one point you said there was no correlation and that was obviously false. There is a correlation in the actual numbers which makes for a useful projection tool.

As far as comparing projections to actual data, I've done that exercise before with past seasons. Typically overall win totals for teams end up being somewhere between 65% and 80% depending on who's projections you're looking at. ZiPS is usually one of the highest of those projections year after year but I'm certain there are examples where they were horribly wrong. If there was a perfect tool we would all be rich and Vegas would be bankrupt.

One of the things that has made baseball more fun over the last 20 years is the fact that statistics have improved. You can make projections based on data rather than simply by the eye test. Of course those that ignore the eye test do so at their peril as that will always be part of the game. Along with statistics come better ways to project future performance as opposed to going solely on past performance which has led to some of the worst contracts in baseball history. Some teams haven't learned those lessons but some have. Even the ones that have can make mistakes. The great contrast in baseball is it is at once the most statistically inclined of all the sports, in other words there are very few areas of the games where solid data is not available, while at the same time being one of the most unpredictable. Baseball is going to baseball. It's fun to study stats, it's fun to watch the games. Nobody, no matter how informed, knows everything.

 

[TC in Mississippi]

That's not how to measure projections. Due to the fact that teams make trades, certain players get hurt, etc. Also, you'd have to look at a significantly larger sample than one year to argue that predictions (And whatever method) are ineffective.

Correct. Projections are reasonably accurate in a large sample size. People like to say how wrong the projections were on the Nationals in 2015 but they actually finished at 83 wins, 7 short of their Pythagorean total which is what they should won, while the Mets finished with 90 wins 3 over their Pythagorean total. most projections had Washington at about 95 wins last year and they lost significant time from starters and very well could have lost 5 WAR in the process (I'm not going to try to prove that as I don't have unlimited time). So with all that taken into account how wrong were the projections? You could take the Cubs projections of 85 wins by most models and do the same exercise in reverse by showing players, notably Kris Bryant and Addison Russell, far exceeding their projections and looking at their Pythagorean win total of 90 games.

You're absolutely right in that measuring projections is not one of "oh that was wrong" but rather to look at a larger sample size and find out how accurate they actually were. Statistical accuracy and predicting how teams do are not really the same conversation.

 

[brett05]

At one point you said there was no correlation and that was obviously false. There is a correlation in the actual numbers which makes for a useful projection tool.

I do not believe you have shown the correlation. I think a lot more is needed that a player's individualized WAR predicts a win. I even stated that I doubt a game's WAR equals a win of 1.0 in a game expecially since the definition is far from standardized.

As far as comparing projections to actual data, I've done that exercise before with past seasons. Typically overall win totals for teams end up being somewhere between 65% and 80% depending on who's projections you're looking at. ZiPS is usually one of the highest of those projections year after year but I'm certain there are examples where they were horribly wrong. If there was a perfect tool we would all be rich and Vegas would be bankrupt.

We agree.

One of the things that has made baseball more fun over the last 20 years is the fact that statistics have improved. You can make projections based on data rather than simply by the eye test. Of course those that ignore the eye test do so at their peril as that will always be part of the game. Along with statistics come better ways to project future performance as opposed to going solely on past performance which has led to some of the worst contracts in baseball history. Some teams haven't learned those lessons but some have. Even the ones that have can make mistakes. The great contrast in baseball is it is at once the most statistically inclined of all the sports, in other words there are very few areas of the games where solid data is not available, while at the same time being one of the most unpredictable. Baseball is going to baseball. It's fun to study stats, it's fun to watch the games. Nobody, no matter how informed, knows everything.

We agree again

 

[TC in Mississippi]

I do not believe you have shown the correlation. I think a lot more is needed that a player's individualized WAR predicts a win. I even stated that I doubt a game's WAR equals a win of 1.0 in a game expecially since the definition is far from standardized.

Interesting. So you actually disagree with fWAR as a statistic measuring wins, even roughly? In other words if a team has a 2 fWAR guy at 1B and goes out and gets a 5 fWAR guy at that position you don't believe that they have upped their projected win total by approximately 3 games? If that's the case than most of this conversation has been rather pointless. WAR is by no means a perfect statistic, particularly in light of the different results of bWAR and fWAR, but I believe it to be a approximate representation of a player's worth to team and their win totals. Most people at least look at it as at least that. If you don't accept that as a reasonable starting point than I've wasted a lot of time arguing with you.

 

[brett05]

Interesting. So you actually disagree with fWAR as a statistic measuring wins, even roughly? In other words if a team has a 2 fWAR guy at 1B and goes out and gets a 5 fWAR guy at that position you don't believe that they have upped their projected win total by approximately 3 games? If that's the case than most of this conversation has been rather pointless. WAR is by no means a perfect statistic, particularly in light of the different results of bWAR and fWAR, but I believe it to be a approximate representation of a player's worth to team and their win totals. Most people at least look at it as at least that. If you don't accept that as a reasonable starting point than I've wasted a lot of time arguing with you.

1) I hope it was not an argument and if you got heated I aplogize for doing that to you. It was never meant to be anything more than a discussion.
2) I think WAR has a ways to go. A common definition is needed first before one can start to seriously address any adjustments it may need. It's a measurable not a statistic.
3) I think I showed with my get 25 two WAR players to show the correlation just isn't there. Now you have said it correlates but I am not sure about that especially in the projection business.
4) I do see WAR as a way to help try and compare players to each other.
5) WAR is also not good at salary projections. It seems to work only in a limited subset as it relates to free agents.

Overall, I use WAR myself in judging players and projecting things like who is a reasonable HOF candidate. I would not necessarily use it for projections at this time for team wins.

 

[SilenceS]

:cringe:

 

[Parade_Rain]

I'm sure most of you have heard the phrase "New Year. New You." Reading Brett05's posts, the phrase I think of is "New Year. Same Brett05."

 

[brett05]

I'm sure most of you have heard the phrase "New Year. New You." Reading Brett05's posts, the phrase I think of is "New Year. Same Brett05."

sorry that you have to change

 

[Parade_Rain]

sorry that you have to change

Trollin', trollin', trollin'
Brett05 keeps on trollin'
Rawhide

 

[brett05]

Trollin', trollin', trollin'
Brett05 keeps on trollin'
Rawhide

the irony

 

[Parade_Rain]

the irony

It's not ironical. It's factual.

 

[brett05]

It's not ironical. It's factual.

I know the irony is factual. :smh: