Diamond Dope Runs Created

[EBH]

It seams that stolen bases are not included in RC under diamond dope.
Am I wrong and just do not see it.
I believe Bill James includes it in his formula.

In the case of Rickey Henderson in Strat o Matic his diamond dope percentages would be much better for RC.

[Valen]

I think I remember reading that somewhere. The difficulty he has is the same I have had in constructing my own database in the past. How to translate Strato ratings in to SB and CS numbers. Just how many SBs should you credit a guy with a 19 good jump rating in ATG where -2 is probably a bad throwing rating? So if he ignores them in his calculation rather than apply some subjective figure to them I can understand. If he does have some formula for converting the rating to expected SB/CS numbers would love to see it and have an extended discussion about it in a thread some time.

[PotKettleBlack]

The best way to add it in would be to use the RL SB plus CS numbers in place of ratings. That's what DeanTSC uses for his Offense vs Defense article, IIRC.

But, the effects on RC of most base stealers is pretty low. I think Rickey's best year was about 7 runs added.

[PotKettleBlack]

To wit:
Regular:
[img:905d263a59]http://upload.wikimedia.org/math/e/3/b/e3b51e8d2bd76d4e16e3a87e028aa46a.png[/img:905d263a59]

With SB adjustment:
[img:905d263a59]http://upload.wikimedia.org/math/1/e/4/1e4f31c0cce7305ca81dd7a218e6ac62.png[/img:905d263a59]

Rickey Henderson, 1982
Regular formula: 81.59
With SB: 92.02

Henderson, 1983
Reg: 89.42
With SB: 104.6

I understated above, but these are two massively successful steal campaigns, and in a 130 SB season (1982), you're talking <11 RC. In a more subtly successful campaign (1983... 108 but a much better conversion rate), you're talking 15 runs.

And that's a couple of the biggest seasons ever. No one else in the modern era stole over 120 in a season, and Henderson would crack 100 only thrice.

[ADRIANGABRIEL]

You are correct, stolen bases are not factored into Runs Created on Diamond Dope.

[dharmabums]

This doesn't change the overall RC stats reasoning (which is fine), but here's a game in which Cobb created 2 runs through stealing bases. Twice, he hit a single, stole a base, then stole another base and scored on the catcher's error. He was also caught stealing once in that game, and twice more in another game. Probably balances out in these particular games, which is why the RC doesn't show much for stealing in general.

http://fantasygames.sportingnews.com/stratomatic/league/boxscore.html?group_id=139249&g_id=9

[Valen]

Not sure I agree with the calculation that has Rickey's 100+ SB seasons only worth 10-15 runs created. Please note I am not doubting his math, I am doubting the accuracy of the values placed on SBs by the creator of the formula.

But setting that aside the impact of a basestealer like Rickey is highly dependent on how you use him and the lineup around him. For example is he preceeded by base cloggers which will reduce his ability to steal second after a walk. Is he followed by sluggers which would exagerate the cost of being caught stealing? That context makes it impossible for any hard formula to calculate how many runs a player like this will generate.

[PotKettleBlack]

[quote:87d02fa0cd="Valen"]Not sure I agree with the calculation that has Rickey's 100+ SB seasons only worth 10-15 runs created. Please note I am not doubting his math, I am doubting the accuracy of the values placed on SBs by the creator of the formula.

But setting that aside the impact of a basestealer like Rickey is highly dependent on how you use him and the lineup around him. For example is he preceeded by base cloggers which will reduce his ability to steal second after a walk. Is he followed by sluggers which would exagerate the cost of being caught stealing? That context makes it impossible for any hard formula to calculate how many runs a player like this will generate.[/quote:87d02fa0cd]

The math on the value of a SB versus the destructive value of a CS is moderate but a little time intensive. You develop a base-out expectancy chart, either based on theory or some large number of games. Man on first, no outs. Two outs, etc.

Now, your steal value is basically the difference between a man on first with a particular out situation, and a man on second with the same out situation. Your caught steal is the difference between man on first, no outs and bases empty, one out. Or all the variation there. And the out is really destructive to value. So, in 1982, Rickey did steal 130 bases. But he was also caught stealing 46 times, which basically nukes the value of 92 steals. So, you have an excess of 38 steals with a run expectancy increase of ~a third of a run, maybe a little less.

Now, say you want to up the complexity, and factor in catcher throwing errors. Not all that common, but still, tilts the value, because instead of 1st and no outs become 2nd and no outs, it becomes 3rd and no outs. I think we've basically increased the run expectancy from .45 to .95. Still, not that large.

But, I want to add a deleterious aspect of stealing in real baseball that doesn't factor in strat: the effect of stealing on the batter, who has to take pitches to enable the steal. With a base stealer on, pitchers tend to get fastball heavier AND get a little worse because of the stretch (very little worse). So, typically, your guy at the plate when your man is stealing winds up with a runner in scoring position, but in a pitcher's count. If he's a good hitter, this is a bad thing, because we might've expected a better result if he just worked the pitcher and hit rather than gotten behind and had to expand his zone. This is why Tom Tango, et al, like putting the base stealer in the 5-6 hole, rather than at the top. Because it won't mess with your good hitters.

@Dharmabums: Your Cobb game:
First Inning: Cobb singles, steals, steals, error home.
Sisler singles (which probably would have moved Cobb two bases)
Lajoie GB(a) (which moves a non-stealing Cobb home from third)

Third Inning: Cobb singles, steals, steals, error home.
Would not have scored otherwise.

Sixth: Cobb singles, steals, knocked home by Sisler

Eigth: Cobb singles, caught stealing.
Sisler, Lajoie and Speaker single behind him.
If Cobb stays put, you get Frank Baker up in a clutch situation. Instead Joe Jackson ends the inning.

Then Cobb is caught twice more in a game we don't have a link to.

What I'm getting at is that the SB DOES create runs, but only when performed at a high rate (65%+, 70%+ 75%+, depending on whose math you like), and maybe not as many as you think.
We have three runs scored with five Cobb stolen bases in that game. And two errors. And 3 singles by Cobb and a single by Sisler. The additive value of the SB is present, but not large.

[dharmabums]

I think you've got it right PKB. What makes it worse, is that those results were for the AAA Cobb. Runs created from steals doesn't get any better than in that game, for the most part.