Most Valuable Player American League

[Oz27]

3 hours ago, AngelsLakersFan said:

 

I feel like you guys are watching a different Albert than I am. The homeruns are nice. The average w/RISP is nice. In general though, watching him play is almost sad, a reminder that father time is undefeated and no one, no matter how great, can avoid their eventual uselessness. 

Yep. The praise he is been getting is just because of his RBI total which is dark ages nonsense. He has 0.804 runners on base per PA, which is the 28th highest mark since 1970. Just by him not being Jerf Mathis or Brandon Wood, that was always going to lead to a heap of RBI. It's more a product of those in front of him than it is what he has done.

[ScottT]

Pujols is doing very well with runners on. Which also means hes doing very poorly without runners on.  

I think his WPA and similar stats are pretty decent. 

[Stradling]

I praise him because he is getting hits with runners on base.  If he had 108 RBI and did it while hitting .200 with runners on base I could see undervaluing his runs batted in.  But when he is hitting .330 or whatever it is, with runners in scoring position, that means something.  

[kevinb]

14 minutes ago, Stradling said:

I praise him because he is getting hits with runners on base.  If he had 108 RBI and did it while hitting .200 with runners on base I could see undervaluing his runs batted in.  But when he is hitting .330 or whatever it is, with runners in scoring position, that means something.  

But in advanced stats are those numbers real? As in do advanced stats account for a clutch or do they eventually even themselves out to what your hitting with runners on base or with no runners on base? Sorry if that didn't make sense 

[Stradling]

Here is how I look at it.  Since we are talking about his season I use that season as a sample size.  So last year he was poor with runners on base and he was judged by that result along with other results.  This year he should be applauded for hitting .330 with runners in scoring position.  

[tdawg87]

I'm done caring about Pujols not producing on a sabermetric level. At this point 30 HR and 100 RBI's is all I'm hoping for.

[Oz27]

14 minutes ago, Stradling said:

Here is how I look at it.  Since we are talking about his season I use that season as a sample size.  So last year he was poor with runners on base and he was judged by that result along with other results.  This year he should be applauded for hitting .330 with runners in scoring position.  

I'd like to think baseball has evolved to the point where small sample (as in single season) RISP stats should not be taken seriously, whether they're good or bad.

[Stradling]

Oz, that is silly.  If we are talking about the season results of a player then his average with RISP is relevant.  

[Oz27]

4 minutes ago, Stradling said:

Oz, that is silly.  If we are talking about the season results of a player then his average with RISP is relevant.  

I don't agree. It's a random subset of a couple of hundred PAs of that player's season. The results fluctuate wildly from year to year, which tells you it's more random chance than skill. I don't consider it at all when evaluating a player. When evaluating a player's season, we should be looking at the full stats as much as possible and not arbitrary subsets of them.

[stormngt]

4 hours ago, mulwin444 said:

It's Most Valuable Player...not "Best Player on a Playoff Team".  Trout is, without a doubt, the most valuable player as of today's date, by any measure.  It took voters a while to eliminate wins as a necessity in terms of Cy Young,  and it seems it will take a little longer for the older writers to be replaced by voters who understand the value of advanced metrics and why the MVP, like the Cy Young, is not a team award.  

It's most valuable player....not best player....period.

[AngelsLakersFan]

48 minutes ago, tdawg87 said:

I'm done caring about Pujols not producing on a sabermetric level. At this point 30 HR and 100 RBI's is all I'm hoping for.

I generally agree with this, but at this point in the season it's hard to look back and say "cool, Albert might be declining but he did everything we asked him to" when we are looking at a 2018 window and seeing what essentially amounts to 8 consecutive seasons of decline in wRC+ and WAR. He's been dropping about 1 war per year, from above average, to average, to below average, to 1 above replacement level. We keep looking at his production and saying, "if he can just keep this up we'll be OK with it" but every year he slips a little further. At this rate this will be his last productive season as a major leaguer. 

It's nice that he was able to mask the additional decline this year with the great RBI totals, but big picture it doesn't mean much when we finish 30 games out.

[tdawg87]

Well we'd finish 25 games out of he was vintage Pujols.

[AngelsLakersFan]

41 minutes ago, Oz27 said:

I don't agree. It's a random subset of a couple of hundred PAs of that player's season. The results fluctuate wildly from year to year, which tells you it's more random chance than skill. I don't consider it at all when evaluating a player. When evaluating a player's season, we should be looking at the full stats as much as possible and not arbitrary subsets of them.

When you are talking about value though, you have to consider it. If we eventually found out that 1 in 100 players had a gene that made them perform 30% better with RISP it would be ashamed if we didn't give them credit for it.

[ScottT]

Results matter, to some extent. Pujols likely wont keep up his clutchness but he has done it this tear, which is somewhat valuable

[Stradling]

55 minutes ago, Oz27 said:

I don't agree. It's a random subset of a couple of hundred PAs of that player's season. The results fluctuate wildly from year to year, which tells you it's more random chance than skill. I don't consider it at all when evaluating a player. When evaluating a player's season, we should be looking at the full stats as much as possible and not arbitrary subsets of them.

Using that logic Albert is an incredible player. 

[Oz27]

Just now, AngelsLakersFan said:

When you are talking about value though, you have to consider it. If we eventually found out that 1 in 100 players had a gene that made them perform 30% better with RISP it would be ashamed if we didn't give them credit for it.

Yeah, sure. If a hitter had much better RISP stats than their full stats year after year, it would be silly not to give them credit for that. But examples of that are rare at best and non-existent at worst. Single season RISP stats aren't really predictive of how that hitter will be with RISP next year, which is at odds with every stat which is useful for measuring a skill. It is possible for exceptions to exist, but I look at RISP stats as little more than random chance.

[Oz27]

3 minutes ago, Stradling said:

Using that logic Albert is an incredible player. 

I have no idea what your point is.

[disarcina]

if I had a vote for MVP it would be a toss up between Donaldson and Altuve and I'd probably lean towards Altuve.

Altuve is unreal.

Trout is the best player in the game today but his team's games have been meaningless since about mid-June.

[AngelsLakersFan]

35 minutes ago, Oz27 said:

Yeah, sure. If a hitter had much better RISP stats than their full stats year after year, it would be silly not to give them credit for that. But examples of that are rare at best and non-existent at worst. Single season RISP stats aren't really predictive of how that hitter will be with RISP next year, which is at odds with every stat which is useful for measuring a skill. It is possible for exceptions to exist, but I look at RISP stats as little more than random chance.

What are the goals of your stat? There are two kinds of metrics: predictive, and descriptive. WAR is not meant to be predictive, but descriptive. It's an attempt to model actual events on the field, like an accounting measurement in a warehouse. A major issue is that a lot of people do not realize this, and have muddied the formulas with predictive metrics. It's understandable because there are plenty of places where it's impossible to take a purely descriptive stance.

Take defense for example. An out is an out. A descriptive metric would make no other judgements, where as a predictive one would determine difficulty (to predict the likelihood of future performance). The problem with being purely descriptive here is that we cannot reliably know the value of each theoretically unmade play. This is where FG uses UZR to estimate the value of made / missed plays. This is descriptive but based on the result of a pseudo predictive model.

When it comes to hitting with RISP, this is another problem area, where a descriptive stat like WAR would prefer to give Albert credit for his hitting with RISP, but it doesn't, not because it isn't predictable, but because it is impractical to calculate it this way. War has a love / hate relationship with complexity. It wants to include things but it also wants to simplify down in order to appeal to more fans and be more simple to calculate. Being based off raw numbers is exceedingly more simple than forcing it to be calculated on the context of every play. 

[tdawg87]

They should give the MVP to Pujols for the lolz.

[Stradling]

39 minutes ago, Oz27 said:

I have no idea what your point is.

My point is that if you aren't going to judge a players performance on the season based on his actual results of that season then you want to look at his entire body of work.  Albert is an incredible player using that method of thinking.

[Oz27]

11 minutes ago, Stradling said:

My point is that if you aren't going to judge a players performance on the season based on his actual results of that season then you want to look at his entire body of work.  Albert is an incredible player using that method of thinking.

How so? FanGraphs has him at 0.9 WAR and a 111 wRC+, so a little over 10 per cent better than the average hitter and halfway between replacement level and average as an overall player.

[Oz27]

23 minutes ago, AngelsLakersFan said:

What are the goals of your stat? There are two kinds of metrics: predictive, and descriptive. WAR is not meant to be predictive, but descriptive. It's an attempt to model actual events on the field, like an accounting measurement in a warehouse. A major issue is that a lot of people do not realize this, and have muddied the formulas with predictive metrics. It's understandable because there are plenty of places where it's impossible to take a purely descriptive stance.

Take defense for example. An out is an out. A descriptive metric would make no other judgements, where as a predictive one would determine difficulty (to predict the likelihood of future performance). The problem with being purely descriptive here is that we cannot reliably know the value of each theoretically unmade play. This is where FG uses UZR to estimate the value of made / missed plays. This is descriptive but based on the result of a pseudo predictive model.

When it comes to hitting with RISP, this is another problem area, where a descriptive stat like WAR would prefer to give Albert credit for his hitting with RISP, but it doesn't, not because it isn't predictable, but because it is impractical to calculate it this way. War has a love / hate relationship with complexity. It wants to include things but it also wants to simplify down in order to appeal to more fans and be more simple to calculate. Being based off raw numbers is exceedingly more simple than forcing it to be calculated on the context of every play. 

Interesting stuff. While WAR is not meant to be predictive, to a large extent it is. If the only bit of information I told you about Player X was that he was worth 5.0 WAR in 2016, you would say he is likely to be in the 3.5 WAR to 6.5 WAR range in 2017. That is of course true of many other stats, such as OPS+, where the goal is still to be descriptive rather than predictive. But in giving you a pretty accurate picture of how that player has done in that season, it is also predictive to a degree as by far the best way to predict a player's future performance is to look at their past performance.

However, that gets much less true when you look at random subsets within the overall picture of a player's performance. Mike Trout has a .919 OPS after an 0-2 count this year (which btw is one of my favorite stats ever) but before this year he had never been above a .700 OPS in the same situation. Some of that may be due to improved approach but it's also likely that a large amount of that is just random chance. It's a subset of less than 100 PAs and when you take small samples sometimes you get crazy results. Put it this way, I don't look at that stat as a likely indicator that Trout will be really good in the future after an 0-2 count. But if he had an OPS above .800 after an 0-2 count for each of the next three years, at that point you would have to accept that was a legitimate skill and not just luck. I put RISP numbers in a similar category to that. I don't look at it as much more than random chance unless a player consistently performs much better or much worse in those situations than their general performance. By far the best way to judge a player is to look at their overall performance, rather than breaking it down into arbitrary subsets and looking for better or worse numbers.

[disarcina]

why are we talking about Pujols' 2016 stats here?  Isn't this a thread about the MVP award?

that being said, Pujols stats, while not MVP worthy -- have been better than expected for the Halos this year and I'll take that.

actually Pujols stats for most of his Halos' career have been pretty good -- I think if you average his stats here they're something like 25 HRs a year, 75 RBI and probably a .265 average.

he did get off to an awful first two months in a Halos' uni which sort of set the stage for all reporting and stories about him ever since. And, actually, his stats have been pretty good.

He still hits into too many double plays, IMO; his stats have had significant help from having Mike Trout in the same line up hitting right before or after Pujols at times.

But I get kind of tired of the narrative that constantly rags on Pujols........his contract was a big one and for ten years but the Halos knew what they were getting, they knew the back end of the deal would be a time of declining cost-benefit ratio --- and, actually, Pujols has performed well for the Halos and IMO, his play on field -- he never appears to quit hustling, trying t win.

Not sure I'd do the ten year deal again like we did -- but overall - no regrets.  I would not put this contract on list with GMJ, Vernon Wells and Josh Hamilton's.

[disarcina]

went back and ran the numbers --- forgot about his second year with us when he spent a good part of the season on the DL -- this was after coming off the poor start the first year so the press was all over -- 'what a terrible contract this is' stuff.

The first year - he bounced back to  hit 30 HRs and had 105 RBI with a .285 BA -- now- to be fair -- many of us will recall 2012 was one of those years where the team got off to a terrible start in April and by mid-May had a dug ourselves a huge hole to climb out of in order to get back into contention. Halos did play well the second half of the season - sort of late -- and many of the post- May games could be characterized as 'no pressure' type situations....not exactly in the midst of contention, pennant chase.

the next year, 2013 -- was when Pujols spent about half the season or so on the DL and his numbers reflect that.

But overall -- and there's 26 games left or so in thi season, Pujols' not quite five year average for the Halos is almost 29 HRs a season, 95 RBI and a BA of about ,270.

It's not the 35 to 47 HR type years he had with St. Louis but we'll take it. IMO pretty good numbers.