What in the world is OWAR?

MCMLVTopps

So, I was looking at some Red Sox stats a while ago and I see this "OWAR"...it stands for "Offensive Wins Above Replacement. Not only do I not know what that is or means, I can't fathom the value of such a stat, as the numbers were extremely low.

Pls enlighten me on this, I'm clueless on this one. Boy, some of those stat things get very deep, you gotta love numbers to read that stuff !!

Tks

dallasactuary

It means just what it sounds like it means. A player with OWAR of 5 causes his team to win 5 more games than they otherwise would have if they had to replace that player (on offense only) with a "replacement" player (understood to be someone right at the line between minor/major league caliber. Calculating it is ridiculously complicated, and it's got flaws like any other stat.

For context on the magnitude of WAR, consider that the worst team in history (essentially comprised of nothing but "replacement players") won 40 games and the greatest ever about 115, for a difference of 75 wins for the season. So to take a team from the baseline of 40 to the maximum if 115, the 25 players on the roster would have to be, on average, 3 WAR better than a replacement player. And 3 WAR is pretty much the line that defines a "good" season. If you have a team and every one of the 25 players has a good season, you'll win a ton of games. Usually, though, you get 10 or so players that platoon on offense and pitch middle relief and even if they play relatively well will still only get 1 or 2 WAR, and more likely several of them will get 0 or even negative WAR for playing poorly. That leaves 15 players or so to take a 40-50 win team to a 90-100 win team to make the playoffs, and its those 15 players that need to average 3 WAR to get there. The greatest season ever is 14 WAR (Ruth) and the major league leader each season is generally around 9 or 10; anything over 5 is an All-Star type season, and anything over 3 is good.

WAR is a stat that fans often disagree with. "How could Player X only have 5 WAR when I saw him win at least a dozen games for my team?". The primary confusion is a misunderstanding of what it was that actually won those games. If Player X hit a walk-off homerun, he does not get 1 WAR for that since everyone on the team contributed to creating the situation where a walkoff homerun by Player X would win the game. Player X might get the largest share (or he might not), but he'll get some fraction of 1 WAR no matter how he plays in any single game. And fans also tend to forget the dozens, or scores, of games where Player X went 0 for 4 - and earned negative WAR for his less than replacement level effort.

grote15

Difficult to understand or to place in proper context? Yes. Myopic? No.

dallasactuary

<< <i>In other words, it's a myopic stat for myopic people.

"Offensive win." What a joke. >>

Well, in the first place "offensive win" isn't really a thing. OWAR and DWAR both refer to "wins", via offensive and defensive contributions, not "offensive wins" and "defensive wins". Bill James Win Shares method does a better job, but WAR attempts to do the same basic thing: if a team wins 100 games and 40 players were on the roster at some point during the season, for how many wins should each player get credit? Historically, and still today, a pitcher is assigned a "win", but that has been and remains a really stupid and useless stat. Everyone on the team contributes to each win in some way, and since winning is what the game is about, the theoretically "best" stat would be the one that properly determined how many wins each player was actually responsible for.

Players win games with their offense (which includes baserunning) and with their defense (which includes pitching), and as the Mets proved even a team of barely capable major leaguers will win 40 games. From those basic premises, WAR determines for each player how many wins his individual production is expected to produce relative to a barely capable major leaguer. The WAR method does a fair job while Win Shares does a great job, but WAR still does a better job of identifying great players than every standard stat combined. That is, if I know two players' WAR and you know those players' hits, doubles, triples, homers, BA, walks, runs, RBI, stolen bases, HBP, GIDP and anything else I've forgotten, I will know which of the two players was better more often than you will. WAR is a cure for myopia, but ironically myopic people can't see that.