Value of speed...

[SultansOfSquat (Man)]

Guys,

I'd like to propose adding a little more value to players w/ good speed by increasing the value of SBs a bit. While I do believe in the XR formula's valuation of SBs (and CS), I do see that the formula does not really account for the impact of speed on the basepaths beyond SB/CS. Unfortunately, there are no standard stats (in Yahoo anyway) for determining how much a player's baserunning skills contribute to run production apart from SB/CS and others like 2B, 3B and maybe GIDP. So I'm thinking maybe we should increase the value of SB a little, but not too much though.

Maybe making SB = .27 instead of .18 would be reasonably realistic. That would add .09 for each SB, which would add ~3.5 to ~5.5 points for a top speedster w/ 40-60 SBs. The idea would be to credit a speedy/good baserunner an extra base over a mediocre baserunner for every 2 SBs he gains, which seems reasonable to me since such a speedster should easily gain an extra 20 or so bases over a slow running slugger who usually just moves one base at a time.

3.5 to 5.5 points would be a significant gain considering there are really only so many batters that produce 100+ points each year, but it's not too much to throw things off balance. With the current system, very few speedsters make it into the 100+ list. And even w/ the adjustment, we probably still won't see that many light-hitting speedsters make the list.

So what do y'all think about this adjustment?

[prhood]

I understand your reasoning Man but I'm not sure that I agree with it 100%. To me speedsters will make the list if they use their speed effectively which includes singles, doubles and triples as well as Sbs. Willy Taveras is a case in point. His speed produces extra hits and extra bases as well as SBs. If (when?) he learns to take a walk he'll become even more valuable.

I'll go along with the majority on this one.

[The Birds (Davey Johnson)]

I agree with Hounds of Hell. For the reasons he stated, it seems a little artificial to bump up the value of SBs at this point.

[SultansOfSquat (Man)]

The reasoning for the idea was mainly because the formula was intended to measure the performance of a batter in his contributions toward run production. While singles, doubles, triples, SBs, etc. all will factor in a player's speed, they do not unfortunately factor in the impact of speed when he is not at the plate or stealing bases. For instance, the formula credits a player for hitting a sac fly, but yet, it does not credit a player for taking an extra base from 1B to 3B on someone else's single or going from 2B to 3B on a medium flyout or turning someones shallow flyout into a run producing sac fly. Those were the kinds of things I was hoping to credit the speedster although doing it via the SB stat is admittedly a bit "artificial" as pointed out. It is too bad that we don't have a better stat for this as I do think the XR formula undervalues speedsters somewhat.

But anyway, I'm not exactly high on patching up the XR formula in this manner although I would like to see a more thorough valuation of speed so as not to significantly undervalue speedy players.

[The Birds (Davey Johnson)]

I finally read Moneyball, three years after publication. I have to say that it was one of the more enjoyable books that I have ever read.

In the book, Bill James' Runs Created concept was featured as a way to measure the output of offensive players. For pitching, the featured concept was DIPS by Voros McCracken:
www.baseballprospectus.com/article.php?articleid=878

While the offensive scoring in this league is loosely based on Runs Created, was there ever any thought to base the pitching scoring on some variation of DIPS, where the key variables are K, BB, and HR allowed per batters faced?

[SultansOfSquat (Man)]

At the time, I didn't dig into it enough to know about DIPS. RC (or rather XR) was an easy choice since it's been around for so long, but I'll have to find out how DIPS works before I'd want to try using that. That article doesn't show the formula though, and one problem I can forsee is that it might not be a simple linear formula like XR, but might be non-linear like RC, which seems impossible to implement in Yahoo.

Since we don't account for defense on the part of the position players, it seems a bit unnecessary to use DIPS for pitching. But yeah, ideally, we should credit position players for their defense (or lack thereof) and try to separate defense from pitching to get a more realistic model. We should probably also adjust for park effects, etc. too.

OTOH, since this is a dynasty-like keeper league, there's nothing wrong w/ using DIPS to help do your long term predictions/valuations of pitchers in this league even though we don't actually use K, BB and HR for pitching points. ;D Certainly, I'd prefer a pitcher w/ good K/BB ratio than one w/out all else being roughly equal. Same w/ the HR rate (and how the different parks might affect HRs as well).

[Imnotcreative2]

DIPS

www.baseballthinkfactory.org/mccracken/dipsexpl.html

DIPS is an interesting topic. The numbers seem to bear it out but intuitively it makes absolutely no sense so I have a hard time buying into it.

[Imnotcreative2]

As long as we are talking about DIPS (which wouldn't work for this league based on what Yahoo has available) and increasing the value of speed maybe you guys should take a look a EqA.

www.baseballprospectus.com/article.php?articleid=2596

[SultansOfSquat (Man)]

I didn't look closely enough at that particular DIPS formula (or rather algorithm ) to say whether it makes intuitive sense to me, but in general, I think the notion along w/ the particular stats involved makes sense to me. However, I'd probably also want to factor in GB/FB ratio also, if I really want something particularly accurate.

Anyway, I agree that the formula seems impossible to implement in Yahoo (as I had suspected). But then, the EqA formula also seems impossible to do in Yahoo. And at the end of the day, EqA is probably no better than XR for our purposes -- the article itself points out that XR is just as accurate as EqA for 1971-2003 (and probably for the forseeable future as well).

Also, it does not look like EqA accounts for speed the way I was suggesting -- it basically accounts for it the same way XR does, ie. only for the batter of record and the baserunner who's making an official SB attempt, not for the baserunner who does something other than to attempt an SB.

Anyway, perhaps, it helps enough that we use a dedicated CF position, which ends up adding some value to speedsters anyway since most of them are CFs in real life w/ some others also playing relatively demanding positions like SS, 2B and RF. It's not ideal, but it's something anyhow...