Comments on: Willie Randolph This and that about baseball stats. Tue, 16 Jul 2013 17:01:55 +0000 hourly 1 By: John Autin Thu, 24 Mar 2011 20:27:34 +0000 Teddy @29 -- If you're comparing Belanger and Ozzie on defense only, then there's certainly a strong case in your favor. Over their entire careers, Belanger had almost as much defensive WAR as Ozzie (20.9 to 21.6), but he did it in about 40% fewer defensive innings. Belanger had 6 seasons of at least 2 dWAR; Ozzie had only 1 such year.

But when it comes to offense ... well, you know.

By: TapDancingTeddy Tue, 22 Mar 2011 21:58:13 +0000 I've always been a fan of Randolph, and so, I'm glad to see him near the top of any list where the quality of his play is recognized.

As for Mark Belanger, I'm probably the only person I know who thinks he was as good as Ozzie Smith. Nice to see his glove got him close to the top 10.

By: DavidRF Tue, 22 Mar 2011 00:19:17 +0000 @16
Re: Collins and pre-1920.

I just noticed that Collins wouldn't be first going back pre-1920. It would be Ty Cobb. Cobb doesn't make the list above because his SB/HR ratio is too low after 1920. His stolen bases were way down in the 1920s and he hit home runs (when he wanted to) :-).

For his full career though, this SB/HR ratio (897/117) is well over five and he had a whopping 159.4 WAR.

Though it could be that the 1920 cutoff was not because that's where the dailies get cutoff (these are full year stats we're looking at anyways). This is probably a "live-ball" chart. Its likely that Cobb's eight SLG titles would have translated to more HR's in livelier offensive times.

By: Doug Tue, 22 Mar 2011 00:12:31 +0000 @26.

Kds: yes, I grant you that it is the scorer's discretion but, from my observation in today's game (and for the past 30 years or more), it is nearly always scored as a SB. Even when the throw is right there and the fielder just misses the catch, it seems that if the baserunner is even remotely in the vicinity of the bag when the throw comes in, he will get credit for the SB. Almost the only errors that are ever actually charged are for allowing the runner to advance on to third (or home). A runner would seemingly have to stumble and fall half-way to second not to be credited with a SB on a wild throw into center (for example).

However, that's today, and official scoring may well have been different back in Collins day. So, yes, my suggestion that Collins success rate may be inflated because of more errors in his time was based on how steal plays are usually scored today.

Certainly, with a lot more errors to deal with in the old days, it's not inconceivable that different protocols may be evolved on how to interpret and score certain types of plays. And, perhaps, a fielder failing to snag a good throw and apply the tag may have been scored as a CS more often then than it is now. Perhaps because it happened more often and perhaps because runners in heavy wool uniforms and low-tech shoes were, on a whole, probably slower than today and thus more likely to be thrown out by sizable margins. But, I suppose we'll never really know for sure.

By: kds Mon, 21 Mar 2011 20:34:30 +0000 @24/ Doug; When there are errors then a success by the batter/base runner goes into the books as a failure. If the batter reaches on an error he gets an AB but no hit, so it lowers his batting average and OBA. Similarily for stolen bases. If, in the judgement of the official scorer, he "should" have been out but for the error, he gets a CS not a SB.

Suppose there are now 2% errors on SBA that change out to safe. Then we are changing an official SB% from 70% to 72%. Back in Collins' day there may have been 6% errors of that type. If so then 67% in the books would be 73% in actuality, and better than 70% in the books today. (All numbers made up for this exercise.)

By: Lawrence Azrin Mon, 21 Mar 2011 19:45:50 +0000 @24/ Doug - when the offensive levels go down, one-run strategies such as base stealing and sacrifice bunting go up. With fewer runs per game, each individual run is more valuable, which makes it worth the risk trying to steal (or at least the manager thinks so...). In higher offensive periods, there are more baserunners and extra-base hits - in short, there are more oppurtunities overall to score, so there is less need to attempt stolen bases.

"...In a low-run environment, does it really make sense to risk losing a baserunner with only a 2/3 chance of succeeding..."

Yes, and this situation was especially magnified in the dead-ball era, where there were hardly any home runs, and fewer walks and doubles (though more triples). I am not argueing what the optimum strategy _should_ be occording to our current analysis, but what the base-stealing strategy actually has been over baseball history.

By: Doug Mon, 21 Mar 2011 18:29:03 +0000 @19 and Eddie Collins.

"If you consider that there were probably more errors and muffs on stolen base attempts (as there were many more errors overall, 90 - 100 years ago) , the "break-even" point was probably a little bit above 50%."

Lawrence, maybe I'm not following your reasoning properly, but to me more errors and muffs on stolen base attempts means more stolen bases and a higher success rate. In other words, the steal rate on plays with errors/muffs is effectively 100%, so since there were more errors/muffs than today, the old-time steal rate is actually inflated relative to today.

So, my thinking is a 67.3% steal rate for 90-100 years ago is more like a 50%-60% rate today, which no manager would countenance for long in today's game for a guy any siginificant number of steal attempts - such a guy would be getting the stop sign most of the time.

In a low-run environment, does it really make sense to risk losing a baserunner with only a 2/3 chance of succeeding. Certainly the incremental improvement in run probability that that one base obtains should be higher in a low-run environment than in a high-run environment, but is the incremental gain enough to justify a 1/3 chance of reducing that run probability to zero.

I don't know what the run probabilities for the different base/out scenarios were in Collins day. But, if (for example), if the numbers were something like 30% scoring probabilty with runner on 1st and 0 outs, and 50% if runner on 2nd, then attempting to steal with a 2/3 success rate would increase your chances of scoring by 10% or so (33.3% versus 30%). But, if the probabilities were lower (say, 40% to 20% for being at 2nd versus being at 1st), then clearly there would be more incremental value in attempting to steal.

BTW, I'm presuming the RE24 and related stats are normalized in some way to correct for differences in eras. Does anyone know?

By: jiffy Mon, 21 Mar 2011 17:55:13 +0000 Re #22, I imagine a lot of them are like Ozzie- 41 of 79 came in the #8 hole, presumably in front of a pitcher in the NL.

By: Downpuppy Mon, 21 Mar 2011 17:25:23 +0000 Kenny Lofton (622/130<5) has 65.3WAR

I find it hard to picture most of the IBBs on this chart..

By: JDV Mon, 21 Mar 2011 16:40:43 +0000 Vizquel (400 SB; 80 HR). I guess he would be there if it SB greather than OR EQUAL TO 5*HR.