Speaker's 1912 card worth the extra $3.15m?

[PotKettleBlack]

[quote:a0e8b2db97="joekendall"]Does anyone think with the new stats that Bernie will show for defense that we will have enough information to answer these types of questions?[/quote:a0e8b2db97]

A lot of the reason I asked for the ones I asked for was to improve the understanding of the value of defense, at least towards winning via the win shares formulae.

I don't know that Bill James' win share metrics will apply smoothly to SOM. I suspect some of it will not (for one thing, you can be infinitely bad in Bill James' fielding metrics, but the SOM fielder ratings have a cap on how bad you can actually be as they don't apply to every roll).

[kmkravitz]

Stayed with 1912 card. He went 5 for 7, then injured in 3rd game of season, out for 3, comes back, goes 1 for 3, then injured in game he returns, and out for 2 more games. Will have missed 5 of first 9 games of season.

Did not see this coming.

[PotKettleBlack]

[quote:cc16c3f015="kmkravitz"]Stayed with 1912 card. He went 5 for 7, then injured in 3rd game of season, out for 3, comes back, goes 1 for 3, then injured in game he returns, and out for 2 more games. Will have missed 5 of first 9 games of season.

Did not see this coming.[/quote:cc16c3f015]
It happens.

He'll play a lot more games before the season is over.

[MICHAELTARBELL]

[quote:2c8b446b62="joekendall"]The cheaper version isn't a better base stealer.

To break it down (1912 - more expensive card, 1920 - cheaper card):

HBP - 1920
BB (L/R) - 1920/Tie
K - 1920
SI - 1912
DO - 1912
TR (L/R) - 1912/1920
HR - 1912
OBP - 1912
GDP - 1920
Clutch - 1920
BPHR - 1912
SB - 1912
Running - Tie
Defense (Range) - Tie
Defense (Arm) - 1912
Defense (Errors) - 1912

If you view everything else as equal, then I believe the defense would be the tie breaker. If you weight different aspects heavier, then you would have to determine which is best.[/quote:2c8b446b62]

Thanks for the catch Joe, you are correct, the 1912 card is a better base stealing card, so determining factor is arm strength and cost (if $$ saved can upgrade elsewhere) :D

[Stormcrow2012]

PKB said:[quote:0a9db5fa9d]Dean didn't include arm strength (or catcher D) in his article.[/quote:0a9db5fa9d]

This must be because arm strength is so important.


Munich Man said:[quote:0a9db5fa9d]
IMO, base stealing is much less important that a good OF arm rating.[/quote:0a9db5fa9d]

I don't know what I was thinking when I said arm strength does not matter and only saves a couple of runs a year,....at most, you guys are right.

[PotKettleBlack]

[quote:2c73215c61="joekendall"]Max:

You can find what you are looking for on base states here:

http://www.insidethebook.com/ee/index.php/site/comments/the_secret_recipes_of_the_run_expectancy_matrix/[/quote:2c73215c61]

Yeah, I have "The Book" on my bookcase. But the problem with applying linear weights and the run expectancy matrix(REM) to ATG, specifically 100-140-200, is that the linear weights and REM are based on real stats from a period of time put up in a different (read: lower scoring) context.

[PotKettleBlack]

[quote:7e5547c2f9="Stormcrow2012"]PKB said:[quote:7e5547c2f9]Dean didn't include arm strength (or catcher D) in his article.[/quote:7e5547c2f9]

This must be because arm strength is so important.
[/quote:7e5547c2f9]

Sarcasm or cynicism, without a smiley, I cannot tell.

I suspect that arm strength is, as I said above, kind of complicated (and that is understatement... it's kind of complicated in relation to the work in the OvDv2 article, which was pretty complicated to put together in the first place) because the application of arm strength is highly variable (offensive context, runner settings, etc) while the number of plays in the field is less variable (at least as SOM handles fielding).

Catcher D is also complicated, as there are a lot of elements to it. Beyond the standard range and error, there's arm, PB, and T-error. It's a lot to unpack theoretically (the way that the OvD article is theoretical... for that matter, half of "The Book" is lab theory, borne out by observation, or observation, backed by lab theory). And, also context dependent, based on base stealers and pitchers...

[JOSEPHKENDALL]

[quote:45bcf1bfdd="PotKettleBlack"][quote:45bcf1bfdd="joekendall"]Max:

You can find what you are looking for on base states here:

http://www.insidethebook.com/ee/index.php/site/comments/the_secret_recipes_of_the_run_expectancy_matrix/[/quote:45bcf1bfdd]

Yeah, I have "The Book" on my bookcase. But the problem with applying linear weights and the run expectancy matrix(REM) to ATG, specifically 100-140-200, is that the linear weights and REM are based on real stats from a period of time put up in a different (read: lower scoring) context.[/quote:45bcf1bfdd]

If we built a RE Matrix based on the leagues for each specific cap that have finished playing, I believe it would give us a very good idea of where to start for calculating the Arm of OF and C.

[PotKettleBlack]

[quote:ec58118ae7="joekendall"][quote:ec58118ae7="PotKettleBlack"][quote:ec58118ae7="joekendall"]Max:

You can find what you are looking for on base states here:

http://www.insidethebook.com/ee/index.php/site/comments/the_secret_recipes_of_the_run_expectancy_matrix/[/quote:ec58118ae7]

Yeah, I have "The Book" on my bookcase. But the problem with applying linear weights and the run expectancy matrix(REM) to ATG, specifically 100-140-200, is that the linear weights and REM are based on real stats from a period of time put up in a different (read: lower scoring) context.[/quote:ec58118ae7]

If we built a RE Matrix based on the leagues for each specific cap that have finished playing, I believe it would give us a very good idea of where to start for calculating the Arm of OF and C.[/quote:ec58118ae7]

Not a small project.

[coyote303]

[quote:760703852a="PotKettleBlack"][quote:760703852a="AeroDave10"]Great points, PKB.

[quote:760703852a="PotKettleBlack"]And in the ATG100-140-200 run scoring environment, what is the win value of the extra base? Small, in my estimation.[/quote:760703852a]

This statement seems to imply that the ability to steal bases and take extra bases on an OF are relatively similar, which makes sense. Don't know how that affects the Speaker debate, but a very interesting conclusion by PotKettleBlack.[/quote:760703852a]

Well, here's some stuff I know. As run scoring in the real world goes up, the win value of a stolen base goes down. The absolute run value doesn't move that much, but the run price of a win goes up, diluting the value of a stolen base.

The value of first to third is a lot different than a stolen base. Stolen base moves your base-out situation less than a single with a first to third. Let's create a bit of base-out notation, like the boxscores use.
Outs - Runners - Event
0 - 0 - Henderson gets on base
0 - 1 - Henderson steals second
0 - 2
Run expectancy goes up, but not that much. About half of what an out would cost in run expectancy (thus the ~70% number).

First to third goes like this:
0 - 1 - Musial - single
0 - 1,3

Big inning. But how much more value is 0 - 1,3 over 0 - 1,2. That's the question on first to third. I could look this up, but again, as the run scoring in the environment goes up (and it's up in ATG100, moreso in ATG140, and absurdly so in ATG200), the win value of the increase in run expectancy goes down.

I suspect the run expectancy of an individual event goes up with the run scoring environment. Better hitters capitalize on high expectancy situations more often. If 90% of the league is better hitters, then yeah, you'd expect them cash in the runner from third more often, which increases the value of the man on third.

So damn context specific, it's very hard to put a value on arms without a lot of simulation. And since ATG doesn't exist as a standalone computer set... it's everyone's opinion until some better fielding data is available.[/quote:760703852a]

If I understand your sample, you are comparing (one) apple to (two) apples! I agree that a single that advances a base runner to third is better than simply stealing second base. However, you have two hitters reaching base versus one, so it's not a fair comparison.

I could argue they are the same if you have two singles:

batter singles and steals
next batter singles; one base advance only

Batter singles
next batter singles and runner takes third

In both cases, you have runners at first and third.

You could further argue that the stolen base is better because it eliminated the chance of grounding into a double play.