Back to Boards

WAR


WAR

Hey Guy - Just noticed the new field in the create a player screen. Do you have an update on this coming or already out there? Thanks.

Re: WAR

Yes, was also going to post, noticed it appear today, oWAR and total WAR.. Curious as well.

Re: Re: WAR

They are not in use yet, but I'm getting them loaded in case I need them... and the main reason is, the Rank calculation is a huge resource suck on the website. Rank has to re-calculate for every league any time anything changes with the players in the league, and just to be safe, it re-calculates before games are played as well. So in case it turns out I do need to eventually move away from Rank, this seems like the next best option... oWAR would give people a pretty good idea in "one number" how to compare players offensively, as does WAR for pitchers, but honestly I haven't done enough analysis to know if it will work... also, there would be the challenge of creating fictional WAR for made-up players... if you're curious, you can try the tool under Bulk Update, and then Edit Players to see what WAR was created for them.... this tool is doing a look-up of real-life players and trying to select a similar WAR.

Again, this is all experimental and not in use in any way at this point.

Re: WAR

I would personally LOVE to see rank go away. I understand that it helps the new guys but I feel like it takes some of the challenge out of the game just my personal opinion. Keep up the good work!!!

Re: WAR

I would keep rank

Re: Re: WAR

Well, there's alot baked into WAR that makes it difficult to translate here. Each player has their own unique:
*Position adjustment based on innings played
*Replacement Runs added based on their number of PA's.
*Park Adjustment based on Park Factor of home Park and PA's, and SP gets even more complex, because the calculations now factor for better pitchers "creating their own run environment".

For WAR to mean anything within any League, you need the value of a theoretical replacement player. If you have different values of replacement player across different leagues, you want a number that translates the same everywhere, and WAR would do the opposite of that. What a 7 win player is in one League wouldnt be what a 7 win player was in another League.

Re: WAR

Agreed Jay that’s what makes it pretty hard to apply to these leagues. Even though Rank takes a fair amount of processing power to calculate, I’m not sure there’s a way around it. I know long timers wouldn’t miss it, but it’s used in a lot of areas now including CPU logic when it drafts for teams, etc.

Re: Re: WAR

I agree with Jay here. I don't see a way to calculate a meaningful WAR metric that wouldn't require the same recalculations rank needs, because it all comes down to a comparison against an estimate of the best players not good enough to make a roster in a given league.

There are certainly ways of making rank more accurate or translating it to a different scale, but I don't think there's going to be a way to cut out the comparisons between all players within the league and still get a number useful for salaries, auto drafts, etc.

Re: WAR

wow, you guys are all way too smart for me.

Re: Re: WAR

War is still a debatable statistic... Don't get me wrong, I like that stat. It is very relevant when comparing players.
But couple of formula exists for that stat.

Guy can even come up with one of it's own that can make some sense...

Who is the replacement player?

Who can acurately define players importance vs position he his playing?... for me relievers and great designated hitters are way too neglected by that stats. Positional adjustment are a bit arbitrary..

WAR can't really define accurately if a hitter is hitting leadoff and see tons of fastballs, vs a hitter beign protected by a great hitter in the lineup, being constantly pitched around the plate by pitchers, is role in the lineup, etc...

War does not consider a grand slam in a 14 to 6 blowout loss, vs a clutch base hit vs a great closer to win a game in Extra Innings.... Ability to handle the pressure, and come up big in pressure situation, stacking RBI's when needed...
A guy like Edwin Encarnacion comes to mind for me... War never gives him too much credit, but that guy would always come up big when runners on bases...

How about the ability for hitters to stand up againts great pitching... some players have a tendency to stack their stats up againts below average pitching... where others will elevate their game againts great opposition...WAR doesn't recognize that...

So many defensive metrics are up there... some Dwar stats are destroying some players that were perceived as the best of the best at their postions defensively... JT Snow being the perfect poster boy of this innacuracy... he was a solid defender that Dwar stats are presenting him as worst than Mo Vaughn in many seasons he played...
Shifts, handness of the pitching rotation, defensive positioning are all neglected by that statistic

So is the innacuracy of establishing a true impact on a baseball game out of a great defensive catcher....

For me WAR is a great guideline, but there is some give and take to be made about it...

Re: WAR

Actually, it was big stat nerd news a few years ago, Bbref.com and Fangraphs agreed upon a set replacement player value. And alot goes into defensive metrics nowadays, and catcher data includes pitch framing in some metrics. As far as calculations for seasons prior to having a great deal of data, its total zone runs, which have a number of ways of counting defensive value based on hitting tendencies, handedness of pitcher, as well simple putouts and assists. Todays dWAR is NOT the dWAR thats retroactively tallied for all seasons prior to well, again, different stats popped up at difgerent times, but the first wave came in 2002 and beyond...

Re: Re: WAR

Dwar is fairly more accurate nowaday.. but nothing still defines the ability of a catcher to call a great game, keep the pitching staff composed on the baseball mound... it's impact is massive on a teams success... you can still have a below average arm, an average defensive skill set... but. be a master at handling your pitchers, calling a great game, managing a ball game... this is just not measurable...but it probably have one of the biggest impact on a baseball game... even more than any gold glover in the field...

Re: WAR

Believe pitcher ERA for innings catcher appears is baked into UZR/DRS. But as a lifelong fan/amateur/coach/all things lover of the game, I completely get what you're saying.

Re: WAR

Instead of a constant calculation of "percentile" which is what the ranks are giving us... (which is why they have to be constantly updated) why not set static values on the different factors...

Batting average of .100 is a value of 1. (with a min of 1)
Batting average of .400 is a value of 100. (with a max of 100)

Then repeat for the other two (I think) stats you use.

Then it's a static number... only updated when underlying real stats are updated, and then only the players that are changed. (Or everyone at progressions, but it's just a one time quick calculation... not ranking and all of the necessary calcs for that.)

You won't have the even distribution across 1-100... but I'm not sure that matters. It would still work for comparison purposes, which is what most people use it for. Drafting, trades, etc.

Re: WAR

Ritz yeah that is worth considering. Thanks.

Re: Re: WAR

Almost like linear weights....

Re: WAR

Yeah, linear weights. That's a good scientific sounding name!

It would also actually be easier to combine the rank with the dW number for drafting.

A 70 rank with a +1 dW might count as an 80-rank in my book.
A 70 rank with a -1 dW might count as a 60-rank in my book.

I guess I could do that now, instead of winging it... but I don't think you lose much (or anything) without the even distribution of 1-100. I dunno... might be worth a shot to see what it looks like...

Re: Re: WAR

From "The Hidden Game of Baseball", one of my favorite reference books.

- The Sheik