I know WAR is far from a perfect stat, but still really crazy that through 40 games he is on pace to beat Babe Ruth's 1923 season for the best WAR season ever. I feel like he doesn't get as much praise as he deserves.
Mike Trout On Pace For Best War Season Ever - (
New Window )
Trout has been in a slump lately. He’s still a better player than Betts. You expect him to have better numbers than Betts by the end of the year.
But that has been typical of Trout's defense this season. He isn't routinely stealing home runs like he did as a rookie, but according to Sports Info Solutions, he has been well-positioned to take away hits such as that one from Wolters. That catch was worth a little more than half a run, according to SIS
There is no way to tell which fielders saved runs even 20 years ago. Elevation and velocity have only recently been measured for every hit. Are they going to take grainy footage from the 20's where guys look like they circled the bases in .5 seconds and make assumptions?
Quote:
May 8: 0-for-3. In the sixth inning, with two men on, the Rockies' Tony Wolters hit a line drive 103 mph into center field. Based on the elevation and velocity, it's a hit 78 percent of the time, according to Statcast data, but Trout was well-positioned, got a good break, caught it running in and doubled up Gerardo Parra at second base. It doesn't look extraordinary:
But that has been typical of Trout's defense this season. He isn't routinely stealing home runs like he did as a rookie, but according to Sports Info Solutions, he has been well-positioned to take away hits such as that one from Wolters. That catch was worth a little more than half a run, according to SIS
There is no way to tell which fielders saved runs even 20 years ago. Elevation and velocity have only recently been measured for every hit. Are they going to take grainy footage from the 20's where guys look like they circled the bases in .5 seconds and make assumptions?
This is the thing. There is a lot more defensive statistical info available now and it's better to compare current players. But his offensive stats are not overwhelming like, say, Babe Ruth's were. Can you really compare their defensive abilities? Or Willie Mays?
So, if Trout sets the WAR record, it's valid for current players, I'd say. But it gets a real asterisk for "all time", at least from me.
Range, DRS, DEF, all of them seem to be inexact.
but...I was an admitted laggard in the acceptance of advanced stats anyway, so maybe I'm just slow to accept it, but I still don't buy the defensive analytics.
Just look at the variation players have from season to season.
One year Alfonso Soriano was the best LF in the NL. I watched quite a few of his games, the guy was a circus out there.
It's entirely possible I don't understand DEF, but how can Soriano fluctuate this much.
2006 LFers:
2007 LFers:
Range, DRS, DEF, all of them seem to be inexact.
but...I was an admitted laggard in the acceptance of advanced stats anyway, so maybe I'm just slow to accept it, but I still don't buy the defensive analytics.
Agree. The defensive stats are a lot more circumspect and then get weighted too heavily in the overall WAR. If you look up SS for a bunch of years Andrleton Simmons had 1 of the lowest total "out of zone" plays, which is completely insane to anyone who has ever watched get to multiple balls nobody else gets to literally every game.
There are a couple reasons. First, defense. Betts does have a (slightly) higher offensive WAR, but Trout has a materially higher defensive WAR. One can agree or disagree with defensive ratings, but that's a big part. Second, "playing time". WAR is a counting stat, and Trout has had more opportunities than Betts. He has almost 10% more plate appearances, and almost 15% more innings in the outfield. That adds up when what you're measuring is aggregate value added.
SFGFNCGiantsFan : 9:49 am : link : reply
Great player who toils away for a team nobody cares about even in their own town.
The Angels are routinely in the top 10 of attendance in the MLB
Quote:
...
SFGFNCGiantsFan : 9:49 am : link : reply
Great player who toils away for a team nobody cares about even in their own town.
The Angels are routinely in the top 10 of attendance in the MLB
I've been to countless Angels games. There's usually a ton of fans from the opposing team.
What you posted about Soriano's defense in 2006 vs. 2007 shows some of the variance in defensive metrics. But for a long time, analysts have said that you need three years of defensive data to have the equivalent of one year of offensive data due to sample size issues (not sure if that's still the rule of thumb).
Also, some players just have really bad years in the field compared with at the plate. They misplay a few balls, take some bad routes, don't get great opportunities to make them up, etc. Sometimes they have bad luck and deal with a higher number of hard hit, tailing line-drives and fewer cans of corn.
No one thinks that offensive metrics are wrong when a player has a great year followed by a terrible year.
Take a look at Bryce Harper in 2015 and 2016 where he had a similar number of plate appearances.
2015: .330/.460/.649. His wRC+ was 197 and hit 42 HRs.
2016: .243/.373/.441. His wRC+ was 111 and hit 24 HRs.
There is a lot of variance on offense as well, but we are so used to offensive statistics and measurements that we accept them as more reliable due to familiarity.
I only ever questioned in my post the defensive part of WAR. offense is what it is, the stats, even the advanced ones, are based on data.
I don't believe defense works that way. It's too variable and not controllable the way offensive stats are.
To me assigned a defensive stat to a baseball player is almost like a PFF rating in football.
opinion.
Just where I am right now on this. As the stats evolve, I expect my stance may too.
I only ever questioned in my post the defensive part of WAR. offense is what it is, the stats, even the advanced ones, are based on data.
I don't believe defense works that way. It's too variable and not controllable the way offensive stats are.
To me assigned a defensive stat to a baseball player is almost like a PFF rating in football.
opinion.
Just where I am right now on this. As the stats evolve, I expect my stance may too.
My point is that offense can have incredible fluctuation as well as it did with Harper's 2015 and 2016 seasons. He went from the 7th best season by wRC+ of the last 50 years (4 were Bonds in SF) in 2015 to a completely mediocre 13th/21 best qualified RFs in 2016. That is a monstrous difference in just one season.
Both offense and defense are based on data, offense just has more buckets that are familiar to us and that we accept as being "objective." But if you stop and think about it, offense has as ton of variability as well. Just look at players' BABIP fluctuation from year to year. Defense has more variability but it's worth contextualizing the variability of offense as well.
Quote:
I won't argue with any of that, other than Soriano went from literally the best fielding LF in baseball to one of the worst in one season.
I only ever questioned in my post the defensive part of WAR. offense is what it is, the stats, even the advanced ones, are based on data.
I don't believe defense works that way. It's too variable and not controllable the way offensive stats are.
To me assigned a defensive stat to a baseball player is almost like a PFF rating in football.
opinion.
Just where I am right now on this. As the stats evolve, I expect my stance may too.
My point is that offense can have incredible fluctuation as well as it did with Harper's 2015 and 2016 seasons. He went from the 7th best season by wRC+ of the last 50 years (4 were Bonds in SF) in 2015 to a completely mediocre 13th/21 best qualified RFs in 2016. That is a monstrous difference in just one season.
Both offense and defense are based on data, offense just has more buckets that are familiar to us and that we accept as being "objective." But if you stop and think about it, offense has as ton of variability as well. Just look at players' BABIP fluctuation from year to year. Defense has more variability but it's worth contextualizing the variability of offense as well.
Agree with most of this.
only I feel liek all the offensive stats or most IMO can be normalized or adjusted to a control. BABIP over time almost always normalizes, even if it spikes one entire season for a specific player I'd expect it to even out eventually.
Some stat geeks can look at a players babip and batting average and tell you within 20 points what the players batting average will be the next season, and man, they're right very often. Show me someone predict a players DEF. Who could have predicted Soriano's variance? And there are a lot of examples like him.
Defense is subjective IMO. Should the player have made the play, were they in a shift, did they say someone else had responsibility for that area, was it late in the game and they were guarding the line for extra bases, was it windy, sunny, what else played a part in the play. Someone physically has to watch the game and make a judgment call.
Offense the numbers usually tell the story.
IOW the eye test IMO is more meaningful to defense evaluation than a stat.
Their model requires players to be rated every play and the reviewer has to make assumptions on what each player's responsibilities were. Not only have they been shown time and again to have subpar reviewers, those reviewers are making judgements on if a player executed the correct plan.
Without knowing that plan!