How do you evaluate a player?

[Tank]

Got to thinking about this after looking at the Starlin Castro thread and was wondering what stats should/would a modern day GM look at in evaluating a player and whether or not to pursue them. Long gone are the days of avg/hr/rbi (unless you're a HOF voter it seems). There are arguments about the value of counting stats vs. complicated formulas, and SABR guys vs. old school. So I'll ask the question here . . .

 

If you were a GM of a real team, what stats would you look at to evaluate a player in determining whether or not to trade for/sign/keep/dump them. Please explain why you'd pick whatever stats you would use, if you don't mind.

[QuinlansMinion]

Theres only one thing to look at and thats OBP.  Unfortunately, this franchise doesnt seem to care about it ever since Disney sold the franchise.  

[nikkachez]

I look at their weaknesses first, see what things they do (positively), and see if it outweighs their glaring weaknesses. I want a team built through the draft and heavily involved in the international market (Caribbean/South America and Asia), a blend of a high-contact and "pesky" hitters to go along with the big boppers (usually the guys that strikeout). I want every position player to at least be close to average defensively (obviously being above average is a plus), I'm a firm believer pitching and defense will win you games. Position players I usually look at OBP, OPS+ or WRC+, potential home/road splits issues (Rockies, Reds, Rangers, etc), I try to determine their fit on a club. I know that's not the best way of looking at talent, but I do think if you have a strong enough nucleus, you can afford to be a little choosy at some positions. A lockdown bullpen is a must nowadays. 

 

Strong defense, deep starting pitching staff that relies heavily on groundballs and strikeouts who limit their walks (weak contact in general now with Statcast), a three-headed monster in the pen supported by a strong support cast of a mix of other types of relievers (sub-marine pitchers, other deception heavy pitchers, sinker pitchers, high-velocity, etc). That blend of contact hitters, guys who get on base, guys who hit for all forms of power (homeruns and other extra base hits), guys who can run. A versatile bench and guys who can play a number of positions in general. 

 

I also don't really like the idea that certain positions have to have a certain type of player. "Well it's third base, so he has to hit for power!" Do not care, if he plays defense and helps in other facets of the game, I'll listen on him. As long as you get power in other areas, even if it's catcher, second base, etc. 

 

I think you have to look at everything from a fresh perspective. I love statistics and the value they provide, but I try and keep a toe in the "old school" mentality as well. I love the way Kansas City, the Yankees, Cardinals, and Pirates make moves and run an org. Those are the ones I want the Angels to model themselves after. 

[Mark68]

As far as hitting stats, I tend to look at OBP, Slg, & Iso. Iso (Isolated power) measures power production, not just homers, but all extra base hits.

 

 

Iso = Slg - Avg

[Docwaukee]

Offense:

wOBA, wRC+, OPS+, oWAR

Home/road and left/right splits of the above

k rate,  #pitches/pa, swing out of zone rate, LD rate.  

 

Base running:

BsR

SB rate

 

Defense:

Defensive runs saved

 

Other:

Douche factor

# games played per year

 

So here is my ultimate sabr response.  I would try to incorporate as many factors as possible into a stat that has the strongest r correlate to wins or runs.  Including park and league factors, proprietary metrics (defense in particular).  I would assign a value.  Then I would adjust that value relative to non-measurable things that could still be of value like makeup, work ethic, likability, etc.  

 

I would also try to create a risk assessment formula that adjusts for age, likelihood of sustainable performance (ex - defense after the age of 30 etc), and whatever other data they have available and work that in to the formula relative to the stat I have created.  

 

Then I would put it all together, take it to the owner and wait for him to tell me to pursue the guy who impressed him during the home run derby last year.  

[Ordos]

Usually, I just let the know-it-alls take care of it for me.

[m0nkey]

Physical attractiveness 

[Second Base]

Well, I think I'd start with need and budget. 

[tomsred]

..

 

Well, I think I'd start with need and budget. 

 

.....and end with AngelsWin screening.

[Tank]

Offense:

wOBA, wRC+, OPS+, oWAR

Home/road and left/right splits of the above

k rate,  #pitches/pa, swing out of zone rate, LD rate.  

 

Base running:

BsR

SB rate

 

Defense:

Defensive runs saved

 

Other:

Douche factor

# games played per year

 

So here is my ultimate sabr response.  I would try to incorporate as many factors as possible into a stat that has the strongest r correlate to wins or runs.  Including park and league factors, proprietary metrics (defense in particular).  I would assign a value.  Then I would adjust that value relative to non-measurable things that could still be of value like makeup, work ethic, likability, etc.  

 

I would also try to create a risk assessment formula that adjusts for age, likelihood of sustainable performance (ex - defense after the age of 30 etc), and whatever other data they have available and work that in to the formula relative to the stat I have created.  

 

Then I would put it all together, take it to the owner and wait for him to tell me to pursue the guy who impressed him during the home run derby last year.  

Wow. I can't imagine the amount of time this would take, but if you're going to do this right you'd better be willing to put in the work. Thanks, doc.

[halomatt]

Got to thinking about this after looking at the Starlin Castro thread and was wondering what stats should/would a modern day GM look at in evaluating a player and whether or not to pursue them. Long gone are the days of avg/hr/rbi (unless you're a HOF voter it seems). There are arguments about the value of counting stats vs. complicated formulas, and SABR guys vs. old school. So I'll ask the question here . . .

If you were a GM of a real team, what stats would you look at to evaluate a player in determining whether or not to trade for/sign/keep/dump them. Please explain why you'd pick whatever stats you would use, if you don't mind.

GRIT

[Tank]

I look at their weaknesses first, see what things they do (positively), and see if it outweighs their glaring weaknesses. I want a team built through the draft and heavily involved in the international market (Caribbean/South America and Asia), a blend of a high-contact and "pesky" hitters to go along with the big boppers (usually the guys that strikeout). I want every position player to at least be close to average defensively (obviously being above average is a plus), I'm a firm believer pitching and defense will win you games. Position players I usually look at OBP, OPS+ or WRC+, potential home/road splits issues (Rockies, Reds, Rangers, etc), I try to determine their fit on a club. I know that's not the best way of looking at talent, but I do think if you have a strong enough nucleus, you can afford to be a little choosy at some positions. A lockdown bullpen is a must nowadays. 

 

Strong defense, deep starting pitching staff that relies heavily on groundballs and strikeouts who limit their walks (weak contact in general now with Statcast), a three-headed monster in the pen supported by a strong support cast of a mix of other types of relievers (sub-marine pitchers, other deception heavy pitchers, sinker pitchers, high-velocity, etc). That blend of contact hitters, guys who get on base, guys who hit for all forms of power (homeruns and other extra base hits), guys who can run. A versatile bench and guys who can play a number of positions in general. 

 

I also don't really like the idea that certain positions have to have a certain type of player. "Well it's third base, so he has to hit for power!" Do not care, if he plays defense and helps in other facets of the game, I'll listen on him. As long as you get power in other areas, even if it's catcher, second base, etc. 

 

I think you have to look at everything from a fresh perspective. I love statistics and the value they provide, but I try and keep a toe in the "old school" mentality as well. I love the way Kansas City, the Yankees, Cardinals, and Pirates make moves and run an org. Those are the ones I want the Angels to model themselves after.

Excellent answer. Thanks. I'm envisioning a Scott Boras kind of notebook for each player.

[Hollyw00d]

Its also important to look at quality of contact, meaning softly hit, medium hit, and hard hit balls. Sometimes those percentages can show a trend from year to year

[Blarg]

Anyone that has a name that brings up the image of a Marxist/Communist leader is scratched from the list immediately.

[ettin]

Wow. I can't imagine the amount of time this would take, but if you're going to do this right you'd better be willing to put in the work. Thanks, doc.

 

Doc basically touched on a lot of it so no need to repeat what he covered and I agree with. The whole process of examining stats combined with actual scouting of the player(s) including getting to know them for who they are as people plays an important part too.

 

Case in point I read the Simmons article someone posted here or on another site and it talked about Simmon's incredible work ethic which is very likely an important factor the Angels considered in obtaining him.

 

These teams have really taken traditional scouting methods and have supported those methods with statistical science to help build these "decision systems" that take in all of the different inputs and assign a value to them and then the system combines these values into a pre-set series of formulas that help to rank these players in real-time to determine their preference to the teams.

 

Each team has a different system with different inputs and different formulas. Some are very simple systems others are much more complex.

 

In the end these systems are there to aid in decision making for the GM. The systems help take the emotion out of the decision making process to give more objective evidence to the front office(s). They will never replace humans but examining and blending all of the relevant stats with traditional scouting, psychological evaluations, medical exams, physical body types, etc. provide supporting evidence to the end decision.

[Brent Maguire]

For offense, I look at wRC+, OBP, ISO and K%. You can tell a lot about a guy by looking at these ones, specifically wRC+. Too many people evaluate based on raw AVG/OBP/SLG without accounting for the environment a player is in. I like looking deeper as well like looking at contact%, swing%, etc. Another useful tool I like is BABIP. It's a great way to see if a guy had a fluke performance or if he's due for some good luck going forward. 

 

From the scouting side of things, I want a guy who controls the strike zone. Raw power, bat speed, barreling the baseball, etc. are big but you need guys who have a plan at the plate. Many guys can get away with big power numbers in the minors but the ones that have poor BB/K ratios almost always struggle in the majors. A good balance of OBP and SLG in the lineup is great but I would want a roster constructed of guys who can give you quality at bats and aren't prone to long slumps. 

 

Defensively, defensive runs saved is my go to choice. UZR is a good choice as well but DRS seems to be a little more consistent. Errors aren't useful to look at without context. A lot of very good defenders have higher error totals due to them just getting to more baseballs. Range is the single most important aspect of defense and you can gauge that with DRS and UZR. 

 

1st step quickness is a big aspect of defense. Many people think the advanced numbers don't do Trout justice but I actually think they're accurate based on my interpretation of Trout. He has elite speed and range but he tends to struggle getting a read off the baseball right away. Go watch the Kiermaiers, Cains, Lagares' of the world and you'll see what I'm talking about. This applies at all positions. We'll all realize how good Simmons is when we see his 1st step quickness in an Angels uniform. A strong throwing arm is also an obvious plus but some guys can get away with it if their range, 1st step quickness and lateral movement can make up for it.

 

On the pitching side of things, I like looking at K%, BB%, GB%, FIP and xFIP. All of those are the best barometers to see what a pitcher is actually doing and what you can expect going forward. I still use ERA and it's obviously useful but often times it's not telling of what a pitcher is actually doing. People say that a pitcher's job is to not allow runs. I don't necessarily look at it that way. I think a pitcher's job is to not allow contact(K%), not give guys free bases(BB%) and if they are allowing contact, for that contact to be weaker and on the ground(GB%). There are a handful of pitchers who beat their peripherals over the course of a career but using ERA to gauge a player on a 1 year or even 2 year sample can be misleading. FIP and xFIP are actually more accurate than ERA for projecting future ERA performance. 

 

Nowadays, it seems like every pitcher throws 95+ so finding big arms is ideal but also not a necessity. I love command guys who have great change ups. Hell, look at what Dallas Keuchel has done the last 2 years. He's a master of putting his stuff where he wants and that's why he's turned into an ace with a huge fastball. Good strike zone control and the ability to manipulate batters can be just as important as velocity. 

Edited November 24, 2015 by Angels_Baseball

[Inside Pitch]

Personally, I've always looked at the tools and then look for the stats to back up what my eyes say I'm seeing.   I don't bother really with non predictive stats or any raw counting stat.  I put a lot of weight in the player's age and for trends in their predictive stats.  O-zone, nB% kS% LS%, all stats that point towards their grasp of the strike zone.  pFB% and LD%  -- stats that indicate the sort of contact the player is making and what he's doing with it, hard hit medium hit, soft hit are useful too.   I don't really like guys that fail at run prevention, or lack athleticism.  Not saying a guy needs to run like the wind, but he needs to be able to fit in at more than just 1B and DH.  

 

Pitching wise..  I look at K rate, GB% and their ability to control pitch counts.  Anyone that can command a gimmick pitch or has a well above average change up could potentially come up huge.  HR/9, BB/9...   those are self explanatory.  I think pitching is a much more inexact science than hitting.  Tools play a significantly bigger role than with hitters IMO.   You can't teach velocity.   There's a lot you can do to improve performance, but using Newcomb/Smith as examples, you just can't teach a guy to throw that hard with as little effort as he does.  Nate Smith and his late season fade is a great example of guy who was coached into a spike in velocity and then faded as fatigue set in.

[Docwaukee]

Wow. I can't imagine the amount of time this would take, but if you're going to do this right you'd better be willing to put in the work. Thanks, doc.

It wouldn't take me long at all.  I would have a bunch of guys from MIT compile the data and tell them to email it to me by the time I get to the 9th hole.  

 

Seriously though? My first order of business would be to do a comprehensive analysis of the mlb draft for players taken after the first 20 picks or so.  Or somewhere in the range of when every player left was available to every team at some point.  With all the advanced metrics and data available, you can't tell me that there isn't a better or more accurate way to find talent.

 

The Angels generate 300mil or so in revenue.  Scouts cost 30-60k per person per year.  From my understanding, the Halos might employ 15-20 scouts.  Even on the high end with directors salaries, expenses, and benefits etc, you are maybe at around 2mil per season.  That's probably a pretty crude estimate, but that is Cliff Pennington's salary over the next two seasons.  

 

Every year there are all star players that come out of the rounds other than the first.  It's not anecdotal.  

 

There aren't a ton of them and of course the odds decrease as the rounds get later, but they are there for the taking.  

 

Every team's scouting budget is different.  Every year you see some article about a team trying something new.  Yet the results are similar.  Is it just not possible based on current info?  Maybe, but that sure seems like the easy answer.  

 

I am probably speaking from a point of ignorance in that I am sure with all of the resources of an mlb team that this has been looked at, but it sure doesn't feel like it's been taken to the extreme that it could relative to how important it is.  

[Tyler]

Prospects in person, major leaguers by TV and statistics.

[Docwaukee]

I forgot to make my actual point. 

 

You likely aren't going to grab every guy that ends up a major leaguer, but what is it worth to dramatically increase your odds?  What if by spending an extra 3-4mil per year on scouting and player development you find one additional everyday player that gives you 2-3 WAR over their six years of club control.  At minimum that's 12 WAR or about 15-20mil per season in savings.  Or at the very least, it creates additional organizational currency that can be used to obtain talent for the major league club.  

[Richard]

How many times they've been on Sportscenter top 10 and how good they do for my fantasy team.

[Guest]

I use my gut. 

[Tank]

Personally, I've always looked at the tools and then look for the stats to back up what my eyes say I'm seeing.   I don't bother really with non predictive stats or any raw counting stat.  I put a lot of weight in the player's age and for trends in their predictive stats.  O-zone, nB% kS% LS%, all stats that point towards their grasp of the strike zone.  pFB% and LD%  -- stats that indicate the sort of contact the player is making and what he's doing with it, hard hit medium hit, soft hit are useful too.   I don't really like guys that fail at run prevention, or lack athleticism.  Not saying a guy needs to run like the wind, but he needs to be able to fit in at more than just 1B and DH.

So, RBIs? Cool.

[T.G.]

I look at how many followers they have on twitter,the number and quality of endorsements they have, and the value of their rookie cards according to Beckett.

Or...

I just take IP'S word on a player.

[gotbeer]

If I was GM of a team.  I would hire Bane and ask him what he thinks.