HS calculation for WAR

Interesting question.

I would think wide disparity in talent, small sample size (not enough games) and inconsistant scoring would render calculating WAR a fruitless exercise.

War (for ML players) is wins against replacement.

In other words how many more wins would the player give over a replacement (of a minor leaguer/ bench guy).

There is actually no formula, and position players/pitchers are calculated differently.

I guess you can look it up, but I suppose that you already knew that.

quote:

Originally posted by JMoff:
Interesting question.

I would think wide disparity in talent, small sample size (not enough games) and inconsistant scoring would render calculating WAR a fruitless exercise.

Here’s a partial answer I got in another stats forum, and my answer.

“

quote:

…In practice, I think it's impossible not only because the statistics are suspect but because the lower the level you descend in talent, the more likely the player's performance could be driven by random events (they play far fewer games) and discrepancy in training. (Fir example, I was a pitcher in high school, but I was never given training on handling bunts, making pick-off moves, not telegraphing my pitches, etc.)

I don’t care about suspect statistics, random events or a lower skill level. If the formula is sound, then its sound. It wouldn’t be as valid as if figured for the ML, but it should still be a good metric when compared to all other metrics. I’d run the numbers and let the chips fall where they may, but I don’t see where it would be possible to get the performance of the average player because there’s no place ALL HS players numbers, good or bad can be found, and my understanding is, everything’s based on that average player calculation. I’d just like to see what would happen if there was a way to run the numbers.”

Sorry to cheat and use something already written, but it seems you and he were saying the same thing, and thus my answer would be the same.

As I noted, its my understanding that in order to run the calculations, you need to know what the numbers are for the “average” player. If that’s true, since there’s no place I know of where ALL HS numbers are kept, it seems to me the entire calculation would be impossible. I was thinking though, that it wouldn’t be unreasonable to get all the numbers for a smaller group.

FI, in our league every team does numbers on every game, and keeps them in the same place, so it would be possible to calculate the “average” player in the league, at least for the seasons there are numbers for. I’m also pretty sure all DI schools(large schools) in our section(Ca has 10 sections, we’re in the SJS section) do numbers and keep them in the same place as well, so they could probably be done too.

But even using only those two relatively small groups, there’s a problem. Other than league games, schools generally play teams from different divisions. We play some games against DII, DII, and even a DIV team once in a while. That doesn’t necessarily mean they play a lower caliber of ball, but rather that they’re likely to not have the same depth of talent because of the smaller pool to draw players from.

But even with all the negatives, when one is working with averages, I don’t see how it really matters a whole lot. Am I seeing this thing all wrong? I’m only looking for what it seems is available for the ML players. Maybe its possible to have a localized “CONFLICT“ rather than a “WAR”.

quote:

Originally posted by TPM:
War (for ML players) is wins against replacement.

In other words how many more wins would the player give over a replacement (of a minor leaguer/ bench guy).

There is actually no formula, and position players/pitchers are calculated differently.

I guess you can look it up, but I suppose that you already knew that.

I know what WAR is, although since I don’t dabble in ML statistics, I’m not as familiar as I could be. The trouble with it for the average guy, at least as I understand it, is that there are many different formulas because they look at offense, pitching, and defense and then combine them.

But, again as I understand it, and when I say that its, assumed I could be in error, everything is based on the average player, and that numbers is fairly simple to get for ML players. As you know, if you divide the total number of hits by every ML player, by the total number of at bats for every ML player, you’d have the average ML BA. I don’t know of any way that’s possible to do for HS players.

I don't know because it's based on replacement, and how can you measure that in HS? Who do you replace who with who at that point by using this?

For instance Pujols is like a 9 or 10, do you know of any bench player or milb that could replace his production?

http://www.baseballprojection.com/war/p/pujoa001.htm

It is definetly new school stuff, that's for sure, and my understanding is that teams are actually seriously using this metric,

quote:

Originally posted by TPM:
I don't know because it's based on replacement, and how can you measure that in HS? Who do you replace who with who at that point by using this?

For instance Pujols is like a 9 or 10, do you know of any bench player or milb that could replace his production?

http://www.baseballprojection.com/war/p/pujoa001.htm

It is definetly new school stuff, that's for sure, and my understanding is that teams are actually seriously using this metric,

I assume that if all HS players’ numbers were kept the way all pro players’ numbers are kept, the formulas used would work exactly the same way.

Admitting my lack of a depth of knowledge about WAR, I looked a little closer at the definitions, and came across this.

WAR calculates the total number of wins that any player adds to his team over the course of a season by comparing the player's performance with that of a fictitious replacement. A "replacement player" is assumed to be an average Triple-A callup who might appear in the majors only as replacement for an injured player, and whose hitting/fielding or pitching skills are far below league average.

Trying to parallel HSB with MLB, to me that indicates replacing a V player with the average JV player who gets promoted to the V in some kind of a similar “emergency”.

Again, I have no trouble believing WAR COULD be calculated at the HS level if the numbers were available, but I’m not at all certain about how knowing a HS player’s worth would help anyone. My guess is, college and pro scouts could use it, but the “problems” in the numbers would almost certainly cause more problems than having the WAR computation would solve.

My guess is, WAR is a stat used mostly by fantasy players, and perhaps baseball executives to make decisions about contracts and player movement. I don’t see how HSB would have much of a demand for any of those things.

I only asked the question because WAR in all of its forms is currently the hot stat in the pro baseball world, and I was wondering if it could be applied to HSB.

alll such calculations can be manipulated--- are we going too far with such calculations?

Stats,
If ficticious replacement is the key word than you could possibly formulate it for V for JV.

Yup, it is a hot stat, I understand that front offices are using it to determine value for contracts as well as trade, and movement.

There a need to show some respect
to those older, wiser, and more experienced.

This is both old and new school.

As far a WAR.
At the 30,000 feet level, think of
trying to improve the objective decision making.

Even if you could access all of the HS stats in one place, it wouldn't give as an accurate a calculation as MLB since we all know HS stats are skewed by mommy and daddy scorekeepers and/or coaches. There is no independent standard consistently applied at the HS level.

quote:

Originally posted by TRhit:
alll such calculations can be manipulated---

Yes, stat manipulation happens to some degree at every level, and probably takes place at the HS level more than any other. But I believe that the more “public” the numbers become, the less manipulation is going to be a factor.

When you get right down to it, how are the stats “manipulated”? Of course the most likely way is that the SK doesn’t know the scoring rules, and things get scored incorrectly. The next most likely way it happens, is because of some kind of bias scoring the game. That may be the boys on the bench scoring the game giving their team every benefit of every doubt, or a parent acting as a team scorer having a distinct “tilt” to their scoring. Then you have the “accidental” mistakes, such as data entry errors that don’t get caught.

But I think there’s an undercurrent of something much more sinister when people start debating the validity of HS numbers, and that’s purposeful manipulation in order to make the team or a player look better than he is in reality. IOW, outright cheating. To be sure, that too definitely takes place, but I believe its something that as a percentage happens so little, I doubt it has much impact at all.

quote:

are we going too far with such calculations?

It depends on the calculations you’re talking about, and what they’re being used for.

quote:

Originally posted by Outfielder:
Even if you could access all of the HS stats in one place, it wouldn't give as an accurate a calculation as MLB since we all know HS stats are skewed by mommy and daddy scorekeepers and/or coaches. There is no independent standard consistently applied at the HS level.

Actually, the accuracy of the calculation would be exactly the same as for MLB. What would be different, is how valid the results would be, and that would depend on what they were being used for.

That raises an interesting question. What do you use baseball stats for? I know that a pretty open ended question, but give it a bit of thought, and try to answer it.

do you really believe there are that many true .400 and .500 hitters in high school? I've seen infield pop ups scored as hits. I've also seen kids that have torn it up in HS, get to college and barely hit .200.....you might be able to calculate a stat, but if the data is biased the stat means nothing.

quote:

Originally posted by Outfielder:
do you really believe there are that many true .400 and .500 hitters in high school? I've seen infield pop ups scored as hits. I've also seen kids that have torn it up in HS, get to college and barely hit .200.....you might be able to calculate a stat, but if the data is biased the stat means nothing.

What’s a “true” .400 or .500 hitter? Do you mean would there be as many players hit that high if all scorekeepers were the same quality as ML scorekeepers? If that’s what you mean, my answer is a qualified no. There would definitely be players that hit .400 and maybe even .500, but not as many. In 5 years, we’ve had 1 .400 hitter on our V team, and in the 4 years my son played, there wasn’t even 1 Varsity .400 hitter.

Whether or not the data is biased means very little to me because I don’t even attempt to compare the players on our team with other players, at least not on a 1 to 1 basis. Since I’ve been the only person keeping score for the last 5 years, the measuring stick has been the same for all of our players, and that’s who the coaches are looking at when making judgments. Our coaches don’t make player decisions based on other team’s players.

I also think people have a bad misconception about how many times hitters get on base, when how they got here was mis-scored and thus affected their BA and other numbers. In 5 seasons, our hitters have made 4,919 Plate Appearances, 3,968 of which were Official At Bats, for a team BA of .304.

If every single error was changed to a hit, the team BA would only go up to .359, which is only a 15% increase. So I don’t see how anyone can argue that the reason there are so many .400 hitters in HS is because scorers can’t tell the difference between a hit and an error. Now it may well be that there are scorers who enter sacrifices and ROFCs as hits, and the software doesn’t catch it.

Here’s an example. In the program I use, I force entry of the number of singles, doubles, triples, and HRs, then compute total hits. However, on MaxPreps, which is typical of sites that allow storage of stats, it only asks for the total number of hits. You can put in doubles, triples, and HRs, but hits aren’t a calculated field. So, if someone wanted to, they could enter any number they please, and the numbers would calculate as if nothing was wrong.

What I’m saying is, there are a lot of reasons HS stats aren’t very trustworthy.

MODERATOR NOTE: I deleted about 10 posts in this thread... off-topic arguing between members which detracted from the original discussion. If you have any questions or concerns, please PM me so that we don't push the thread off-topic again.


Thanks,

Julie

I can't buy into high school stats because they are stats in small numbers. Here's a perfect example. I remember the numbers from my senior year because I've used this so often as an example.

The team was mediocre. I had no one batting behind me for protection. A typical game was 1-2 and two walks. But one game I was pitched to and went 5-5. With the short cold weather high school season I finished 19-30 and 30 walks.

The 5-5 game made a huge difference in my batting average when each hit was worth 33 points. I didn't get the ball out of the infield once in those five hits. I beat out three toppers and bunted for two hits (they gave me the line twice). I got 165 points of batting average on garbage hits.

This is why when my kids came through high school I told them to look for five bunt singles per year. Those five hits made them first team all-conference because coaches who only see you play a couple of times a year will look at batting averages before voting. With 50 at bats those five bunt hits were worth 100 points of batting average.

quote:

Originally posted by RJM:
I can't buy into high school stats because they are stats in small numbers. …

So what’s the magic number of at bats a player has to have before his numbers are considered valid? 100, 200, 300? There were 14 Phillies who got to bat last season but had less than 100 at bats, only 12 had 200 and 9 had 300. Pick one of those lower limits, and by your theory, all of those numbers should be stricken from the books and totally discounted because of the “small numbers”.

Heck, why not go all the way and just discount anything prior to whatever magic number you choose because they mean nothing.

HS stats mean very little to college coaches...

The problems with HS stats are many, which have been repeated here many times. First – garbage in garbage out syndrome. Scorekeeping at the HS level is suspect at best. Second there is a wide range of competition in the levels of play, within and outside of regions. In SoCal there are 6 major divisions of play and an additional 6 minor variations within those division. You put a IA player in a 6AA league and the statistics are meaningless. Last year there were equipment variations across the country, and this year these changes will be put in place nationwide. There are also wide variations in fields at the HS level. My son played on fields where he was credited with a HR when he acknowledged after the game it would have been a pop up out at his home field. Some infield’s look like cow pastures and others look like MiLB stadiums. I am not sure what number you need to get to as far as AB’s for statistical relevance, but when you add up the above it does not really matter IMO. It does make for an interesting discussion however.

quote:

Originally posted by Stats4Gnats:

quote:

Originally posted by RJM:
I can't buy into high school stats because they are stats in small numbers. …

So what’s the magic number of at bats a player has to have before his numbers are considered valid? 100, 200, 300? There were 14 Phillies who got to bat last season but had less than 100 at bats, only 12 had 200 and 9 had 300. Pick one of those lower limits, and by your theory, all of those numbers should be stricken from the books and totally discounted because of the “small numbers”.

Heck, why not go all the way and just discount anything prior to whatever magic number you choose because they mean nothing.

Which high school conference to the Phillies play? Most MLB players have established their value over time. MLB versus high school is apples and tubas.

quote:

Originally posted by Outfielder:
HS stats mean very little to college coaches...

High school stats may draw attention. But after that the evaluation comes down to mechanics and potential upside. I've seen kids with mediocre high school stats draw college attention due to physical stature and raw potential. One of my 16U players was recruited in high school by throwing 90+ and being 6'4". Every pitch was an adventure. I used to joke his control was anywhere between the pipes (on the backstop).

WAR simply isn't applicable to HS baseball.

Let me explain, The 'replacement' is nominally defined as (by http://www.insidethebook.com/e...ow_to_calculate_war/ )

WAR is wins above replacement. Replacement is defined very specifically for my purposes: it’s the talent level for which you would pay the minimum salary on the open market, or for which you can obtain at minimal cost in a trade.

For nonpitchers, that level is set at -2.25 wins per 162 games, below the average for that league. Since the same stats of an average AL player is better than the same stats of an average NL player (i.e., the AL is the better league), we have different replacement levels. Those levels are -2.5 wins per 162 games in the AL, and -2.0 wins in the NL.

The positional adjustments are:
+1.0 wins C
+0.5 SS/CF
+0.0 2B/3B
-0.5 LF/RF/PH
-1.0 1B
-1.5 DH

These are needed so that an average fielding 1B is not valued the same as an average fielding SS. The DH and PH should be at -2.0 wins (that is, a poor fielding 1B and a DH have the same fielding+positional value). However, as per The Book, it’s harder to DH, so we give them a 0.5 win boost. The PH gets an extra 1 win boost over the DH, because PHing is much harder.

For pitchers, that level is set at .380 win% for starters and .470 for relievers. The exact same pitcher will perform much better in a relief role than in a starter role. So, you need different baselines. You can read the relevant chapter in The Book for that. As with nonpitchers, you need different baselines in each league: .370, .460 in the AL, and .390, .480 in the NL.

For closers, there is a closer replacement level of .570 win%. Any wins above this level gets multiplied by his Leverage Index (LI). That is, while he is not responsible for the LI he finds himself in, he is responsible for his talent that allows him to take advantage of the extra leverage (sort of like Ozzie Smith gets to play SS and reaps the benefit of all the extra opps). We are only giving credit to the closer for the leverage above the .570 win% level. Credit GuyM for this insight.



--------------------------------------------------------------------------------

This section below is not required to understand the above. It is presented here for the hardcore among you.


--------------------------------------------------------------------------------

To convert wOBA for a hitter into wins: (wOBA - .338) / 1.15 * 700 / 10.5 will give you wins above average. (The .338 is whatever the league average wOBA is, which is EXACTLY equal to whatever the league average OBP is; 1.15 is the relationship between wOBA and runs; 700 is the number of PA per 162 games; 10.5 is the relationship between runs and wins.) Adding in the wins above average at the position plus the positional adjustment gives you wins above average per 162 games. Add in the replacement level, and that gives you WAR per 162 games. Simply multiply that number by the percentage playing time you expect (no more than 90%, typically at 85% for regulars), and you have your WAR.

For starters, you multiply the wins above replacement by the number of full games (expected innings divided by 9). Use 20-24 for your full-time starters. Use 7-9 full games for your relievers.



--------------------------------------------------------------------------------


Ok, so how many wins above replacement are there in AL and NL?

For nonpitchers in AL:
2.5 * 8 * 14
+ 1.0 * 1 * 14
+ 2.0 * 0.1 * 14
= 297 wins.
That’s 2.5 wins for every average AL nonpitcher at the 8 fielding positions, times 8 AL nonpitchers times 14 teams. The second line is the average hitter as a DH. The third line is the average hitter as a PH.

For starters in AL:
(.490-.370) * .66 * 162 * 14 = 180 wins
That’s a .490 win% for the average starter, times 66% of the innings per game, times 162 games times 14 teams.

For relievers in AL:
(.520-.460) * .34 * 162 * 14 = 46 wins

For closers bonus in AL:
(.610-.570) * 1.8 * 8 * 14 = 8 wins
The 8 is the number of full games (72 innings) we give a closer. The 1.8 is his LI.

Adding it up, and we get 531 replacement wins in the AL. Nonpitchers get 56% of that and pitchers get 44%. Of pitchers, starters get 77% and relievers 23%.

Repeating for the NL.

For nonpitchers in NL:
2.0 * 8 * 16
+ 0.5 * 0.05 * 16 (DH)
+ 1.5 * 0.40 * 16 (PH)
= 266 wins.

For starters in NL:
(.490-.390) * .64 * 162 * 16 = 166 wins

For relievers in NL:
(.520-.480) * .36 * 162 * 16 = 37 wins

For closers bonus in NL:
(.610-.570) * 1.8 * 8 * 16 = 9 wins

Adding it up, and we get 478 replacement wins in the NL. Nonpitchers get 56% of that and pitchers get 44%. Of pitchers, starters get 78% and relievers 22%.



--------------------------------------------------------------------------------

As you can see, despite having 2 more teams, there is more talent in the AL, on the order of +53 wins. This was of course manipulated by me with the replacement levels. This would imply about $120MM more payroll (above the minimum) in the AL than the NL. And that’s close to reality.

The total number of replacement wins is 531 in the AL and 478 in the NL for 1009 wins (average of +33.6 wins above replacement; since the average team wins 81, that sets the replacement team at 47.4 wins, or 0.292 win%).

The average AL team is 531/14= +38 wins above replacement (or +.234 wins above replacement per game), while the average NL team is 478/16= +30 wins (or +.184 wins above replacement). Since the replacement level is .292, that gives us the following as the average win levels in each league:
AL: .292 + .234 = .526
NL: .292 + .184 = .476

A head-to-head of a .526 team facing a .476 team would result in a win% of .550 for AL. That’s also the reality of the situation, and the replacement levels for the two leagues was set with this in mind.



--------------------------------------------------------------------------------

The totals by position:
563 wins nonpitchers
346 wins starters
100 wins relievers
1009 wins total

This implies that 56% of your payroll (above the $400K minimum per player) should go to nonpitchers, 34% to starters, and 10% to relievers. For reasons of risk, you might give a bit less to your pitchers.
________________________________________________________________________________________

How the bleep do you create an equivalent level of comparsion for HS players? Define what the average JV player how could be brought up? At what level of play do you define it? The school with 5,000 students or the school with 50 students?

IMHO this stat was invented by MLB agents preparing for arbitration hearings. When the MLB club presents their hour of why 'dude' isn't worth his number. The agent can come back with a vomit of data and the 'substantiated' conclusion, "You won the division by 2 games and my guy's WAR was a +3, without him, you stay home, so he's worth the number we presented."

Totally inapplicable to HS baseball and we should just stop talking about it.

WAR is applicable to Little League.

MoneyBart.

quote:

Originally posted by JMoff:
WAR simply isn't applicable to HS baseball.
…
How the bleep do you create an equivalent level of comparsion for HS players? Define what the average JV player how could be brought up? At what level of play do you define it? The school with 5,000 students or the school with 50 students?

IMHO this stat was invented by MLB agents preparing for arbitration hearings. When the MLB club presents their hour of why 'dude' isn't worth his number. The agent can come back with a vomit of data and the 'substantiated' conclusion, "You won the division by 2 games and my guy's WAR was a +3, without him, you stay home, so he's worth the number we presented."

Totally inapplicable to HS baseball and we should just stop talking about it.

I knew there was someone out there who used his head for something more than a hat rack! Actually, I think it does have an application to fantasy players as well as contract negotiators, but although there’s a lot of fantasy about HS players, there ain’t no fantasy leagues for them!

I believe that if the mechanical problems with doing it for HS players could be overcome, valid numbers could be calculated. The only thing is, they would represent something very different from ML numbers. But that’s what averaging is supposed to do. The trouble is, there’s no application for it that I can see, no matter how valid it could be made.

Its just that I see a lot of pub about WAR now-a-days, and its being treated like the answer to all baseball stats. Unfortunately, there are many ways those stats are used, and none of them have an application in HSB.

Stats4Gnats aka Scorekeeper (who was tossed off the board) has spent the past eight years trying to prove he knows more than high school coaches even though he's never coached. He refuses to grasp stats in small numbers are skewed. He refuses to grasp one excellent or weak game can drastically skew stats when in small numbers.