Archive - player evaluation RSS Feed

Testing the Major League scouts

November 14, 2007 in player evaluation, San Diego Padres, scouting with 0 Comments

A lot of times we (me, at least) like to think that stats are very important when measuring major leaguers. I tend to think that we’re right much of the time. If you put the data in the right hands of people who know what they’re doing, I’ll take numbers over a scout’s opinion on a guy who is a major league vet. And that doesn’t mean scouting is useless in the majors, or any such nonsense. As the brightest executives have been saying for years now, the main idea is to gather all of the data that you can, and then properly weight it to make your decisions. A high school senior — hello scouts. An established major league ball player – the numbers, for the most part, take precedence. However, even when you think you have a guy pretty well pegged with your prediction model, scouting can always provide some additional insight. Did a guy’s bat speed slow down over the course of the year? Is he playing with an injury? Has a prior injury changed his mechanics? The list goes on and on. Another area where scouting is significant is when players are coming back from serious injuries. Numbers from 2-3 years ago are not going to tell you too much about a guy who is a few years older and coming of an injury.

So, with the Padres apparently looking at Matt Clement, their scouts will be put to the test. Clement is going into his age 33 season and he’s pitched just 65.3 innings over the last two years. Coming into 2006, he had a weighted era+ of about 111. Going into 2006, he was probably projected at around 170 innings or so and a slightly above average era. Well, he had all kinds of trouble and had to shut it down on June 14th (and later have shoulder surgery). He posted a 6.61 era and a 43-38 k/bb ratio (the worst of his career) in 65.3 innings.

Like I said, all of that means very little at this point. Any projection of Clement is going to be even more unreliable than a projection of a normal player (and probably worthless, for the most part). I suppose it may be interesting to see how pitchers with similar stuff did coming off of similar injuries, so we’ll keep out eyes out for PECOTA’s projection and similar analysis. But still, any evaluation of Clement is mostly in the hands of scouts. Let’s hope they get it right.

On this market, Clement could be a steal, especially if they can sign him to a low base, incentive oriented deal. As usual, we’ll see what happens.

Should Giles or Kouzmanoff change positions?

November 10, 2007 in fielding, player evaluation, Sabermetrics, San Diego Padres, scouting with 2 Comments

The goal for a defense is to convert batted balls into outs. Therefore, the goal for a manager is to put players in the position where they give you the best chance of converting balls into outs (am I right, people??). If you’ve got a guy whose skill set plays nicely at third, but poorly at short, (in most cases) you probably want to play him at third base. If you do this around the diamond, you should give your team the best possible alignment to play good defense.

It isn’t that easy, of course. Imagine you only have eight fielders for eight spots. If one guy plays better at a different position, but would require you to move another guy to a position where he plays worse, the move is basically pointless. Also, there are other factors like a players willingness to move and his comfort at his position. Playing positions like center field and shortstop generally comes with a bit of swagger, and guys aren’t always ready to believe that it’s time to move down the ladder.

For now, we’ll try to simplify it a bit, and see if there are any Padres that should be changing positions in 2008. To me, two players immediately stand out: Kevin Kouzmanoff and Brian Giles. Kouz stands out because he wasn’t overly impressive at third and (perhaps more importantly) top 3b prospect Chase Headley is starting to breathe down his neck. If, for example, there are two players for two position (Kouz and Headley for 3b and LF), you want to put each guy in the position that best suits his skills. We have some data on Kouz, so we’ll see where he should play (and ignore Headley for now).

The second guy, Giles, comes to mind because he really profiles as a left fielder (decent range, weak arm), rather than a right fielder. The only real difference between those two positions is arm strength (which, in the end, probably isn’t a huge deal), so it probably makes sense for Giles to make the trip over to left. How are we going to do this, though? Numbers aren’t going to help up much, especially if a player has only played at one position in recent years (or only has one year under his belt like Kouz).

Well, thankfully, Tom Tango has been collecting some great data in his Fans’ Scouting Report. Fans are asked to fill out a ballot on each player and rank them in seven skills (instincts, first step, speed, hands, release, arm strength, arm accuracy). Tango publishes each players position-neutral ranking, but in a blog post he also gave weights for each position. That allows us to insert players into other positions to see if they should be moved.

We’ll start with Kevin Kouzmanoff. I don’t think Kouz stands out as a guy who I’d move to the outfield. However, as we’ve mentioned, he wasn’t overly impressive at third and Headley could be ready soon. Adding to that, his arm is a quite wild and that’s something you can hide in left more than you can at third. Anyway, let’s take a look at the scouting data and see if anything interesting comes up. Kouz’s position neutral rating is 40. Here are his ratings at the positions of interest:

third: 43
left: 39
right: 40

It looks like Kouz is better off at third base than in the outfield … but it’s a very slight difference. If needed he could probably move to the outfield and be a similar fielder. If Headley is clearly better at either position then he should play there and Kouz should take the other (note: of course this is a simplified analysis … you’d definitely want to use the opinions of mlb scouts, video, numbers, etc. if you were seriously going about this). Along with this we can also look at Kouz’s comps and see where they play. His top five comps are:

Montero C
Guillen SS (shouldn’t be there … now he’s at first)
Spiezio 3b
Nomar 1b
Nieves C

After a year, I think Kouz is (1) not a great fielder, (2) not a great third basemen, and (3) not playing out of position. I think the Padres will probably stick with Kouz this year at 3b and give Headley another year in the minors. When that debate comes up again, you know where to look ; )

Now onto Brian Giles. Overall, position-neutral Giles gets rated at a 56 by the fans. By position:

right: 56
left: 57

Again, nothing mind blowing. All other things being equal, there’s no reason why Giles can’t scoot over to left. But, in the end, it’s probably not a huge deal. If San Diego thinks they can get a good right fielder over a left fielder, they should probably go that way. If they bring in a guy like Geoff Jenkins, Giles should probably be the one who heads to left. Anyway, let’s look at Giles’ comps: 

Dejesus: CF
Pena: 1b
Phillips: 1b
Bowen: C
Iguchi: 2b

Interesting group, I think (check the rest of his list, as well … some outfielders pop up along with some slower middle infielders). Bowen only had 3 ballots, so I may be inclined to toss him out. By the way, Kouz and Giles each have 15 which seems to be a pretty solid number accordig to Tango (of course, you’d like more … get out there and cast your ballot next year … Screw politics. This is what you should be voting on!)

So, with all of that, we’ve learned nothing. Okay, well, we’ve learned that these two players aren’t extreme enough in their skill sets to make a huge difference where they play. Kouzmanoff could possibly play a corner outfield spot if Headley is much better suited for third. Giles could surely move to left, but he can remain in right without any real concern. We’ll see what happens …. 

I may look at some other players sometime. If you have any suggestions throw them out there. I’m sure there’s a way to get all the ratings real quick at each position, but I’m not very excel/computer savvy. So we’ll go one at at time …

The great debate

November 8, 2007 in Khalil Greene, player evaluation, Sabermetrics, San Diego Padres with 18 Comments

No, it’s not stats vs. scouts. No, it’s not “bashers” vs. “gushers” (whatever the hell that is). It’s not about replacement level.

As The Phantom of the Padres blogosphere has declared, it’s the debate between those who love Khalil Greene and those who don’t. Now,  we’ll discount those who love Khalil because his blond locks (not that their opinion doesn’t count!) and try to look at it from a baseball standpoint. For the heck of it, we’ll try to be objective about it, because that’s kinda the whole point (although it’s hard, if not impossible, to be completely objective). Anyway, first a little background on Greene and why this divide may exist amongst hardcore Padres fans.

  • He makes a ton of outs. It has been drilled into people’s heads that making outs is terrible. And it really is.
  • He plays great defense that is sometimes believed to be overrated.
  • He plays shortstop, which is the position furthest right on the defensive spectrum (not counting catcher).
  • He has very good power for a shortstop, which sometimes may be overlooked.
  • This is not very baseball related, but he has a “strange” personality and may sometimes come of as not caring (I believe that is bs, but this isn’t about me ; ).

Alright, with that out of the way, let’s try to tackle this in three steps. First, let’s estimate his offensive value. Then we’ll try to estimate his defensive value. And finally, we’ll add in a positional adjustment for playing short. After all of that, we’ll check out Tango’s Salary Scale to see approximately how much money he would be worth if he was a free agent.

Offense

For offense I’ll admittedly go the easy route and just use BP’s EqA. I’d prefer using some form of linear weights, but I don’t have that tabulated (although I may sometime soon) and I don’t believe it’s available anywhere. The advantage of EqA is that it’s throws in steals and, more importantly, is park adjusted. Despite being overly complicated, it will work pretty well to estimate his offensive value. According to Tango, you can convert EqA to runs quite easily: PA * (EqA – league avg. EqA):

2005: +0
2006: +1.4
2007: +2.7

Weighting that 5-4-3 it comes out 1.6 runs above average. We’ll call him average on offense (without regarding position).

Defense

The Fans have him rated at 79. The overall average for shortstops is about 59. Assuming .7 runs per point (which is what Tango et al. are assuming, I think), that puts him at about +14 runs on defense this year.

John Dewan’s plus/minus has him at +7 plays this year, which is about 5 runs above average. MGL’s UZR has him at +9 per 150 from 2003-mid2007. We’ll call him +10 runs (0r 1 win) defensively.

Positional adjustment/aging

For position, we’ll tack on the usual +5 for playing short. We don’t need to do much for aging, as Greene is right in his prime at 28.

So, in runs above average (just approximations):

Hitting: +0
Fielding: +10
Position: +5

So, going forward, I think it’s safe to evaluate him as a ~15 runs above average type player. Replacement level is around 20 runs below that for a full season, so he’s about 35 runs above replacement level. Compare that to what JinAz just did, and he’d be about the third best position player on the Reds.

Yet another update: I kind of account for it down below, so it shouldn’t effect the numbers much, but I should mention it … you can’t just add on the 20 runs above replacement, because that’s assuming a full year (I believe). If we project Greene at about 80% of a full season, that means we should add around 16 runs above replacement, instead of the full 20.

Salary Scale

Tango’s Salary Scale attempts to find what a player’s worth on the free agent market. Let’s call Greene a 3.0 WAR player (we’ll take off a little for regression/injuries) just to be conservative. That puts him in line for contracts like these:

1 year/12m
2 years/23m
3 years/33m
4 years/41m

Greene of course is under control until 2010 and will make considerably less than that. Last year, for example, he made about 2.25 mill. 

There are still many things to talk about and discuss with Greene. I mean, for one, saying he isn’t a moneyball player is only based on the idea that moneyball=walks/obp and that is not really true at all. You could argue that his outs certainly impact him negatively as a ball player (they certainly do), but his power/position/fielding make up for it enough to make him into a well above average player.

Still, even acknowledging that, it’s not out of the question that Greene could, at some point, be traded. I don’t think organizations really put anyone out of the question when it comes to trades. If a trade can bring in more value than it gives up, in most all situations the team will probably make the deal. Of course, that’s complicated by many factors such as money, prospects, replacements for the traded player, and so on.

But please don’t mind me and let the debate rage on!

(this was not meant to be a true analysis of Greene — there are many assumptions made, shortcuts taken, etc … just a mere approximation of his value)

Update: Tango has the new Salary Scale up. I was using last year’s, which is based on 4m per marginal win, instead of 4.4m (due to 10% inflation). The numbers are relatively close, but I figured I should let you know, anyway.

Maddux returning

November 5, 2007 in contracts, player evaluation, San Diego Padres with 3 Comments

Good news here. It looks like Greg Maddux is going to be resigned to a one year, $10 million deal. It’s apparently not quite official, but looks to be pretty darn close. According to Tango’s Salary Scale the Padres are valuing Maddux as about a 2.5 WAR (wins above replacement) player. Let’s say Maddux is a 3.90 era true talent pitcher in Petco and replacement level is about 5.00. Project him for 180 innings and that’s about 22 runs above replacement level (or about 2.2 wins). So, they appear to be making a solid move here at least. *those numbers are only rough estimates, of course.

Funny thing,  my estimate wasn’t that bad — ZIPS projects Maddux with a 3.91 era and 200 innings next year. The innings projection seems a bit high to me … if you do a weighted (5-4-3) average of his last 3 years, Maddux comes out at 209 innings. Then you have to factor in aging for a pitcher who is going to be 42 years old. A 9 inning decrease does seem light considering both his age and what we know about pitchers being injury risks. However, as we’ve mentioned before, Maddux is a rare breed as he’s thrown at least 198 innings a year since I was one year old (1988).

Anyway, let’s take a quick glimpse at Maddux’s comps via baseball reference, and see what happened to them in their age 42 seasons. I’ll calculate their 3 year weighted average from 39-41 and then their innings pitched at age 42 (just for the heck of it … ). Here we go:

Don Sutton
3 year: 216
Age 42: 192

Tom Seaver
3 year: 222
Age 42: Retired

Warren Spahn
3 year: 266
Age 42: 260

Gaylord Perry
3 year: 238
Age 42: 151

Pete Alexander
3 year: 234
Age 42: 132

Steve Carlton
3 year: 170
Age 42: 152

Tom Glavine — he’s 42.

Roger Clemens
3 year: 199
Age 42: 211

Eddie Plank
3 year: 223
Age 42: Retired

Bert Blyleven
2 year: 134
Age 42: Retired

Okay, so that leaves us with six guys if we don’t count the guys who retired. I suppose that’s reasonable since they certainly would have racked up innings, but it also shows how old this really is (i.e., many greats are retiring). The six guys that we have left averaged 221 innings with their 3 year weighted averages from 39-41. In their age 42 season, they averaged 183 innings. That’s 83% of their weighted average, so if we apply that to Maddux real quick we get a projection of about 173 innings. I’m not saying that’s more accurate than the 200 innings projection (which was certainly calculated in a more rigorous attempt, I’m sure), but it does show that this age is dangerous for pitchers, even ones who are of Maddux’s caliber.

All in all, though, I like this deal for San Diego.

UPDATE: More on Maddux from Tango (post 16). Tango says thumbs up.

Dan Szymborski (ZIPS creator) also likes the deal. Who doesn’t?

Goodbye Giles, hello Zarate

October 27, 2007 in Padres notes, player evaluation, San Diego Padres with 0 Comments

Marcus Giles was certainly a big disappointment with the Padres. Now, he’s been put on waivers and the Padres owe him 500k to buyout his 4m option. Towers:

“He had a terrific April for us; he was one of our better hitters,” general manager Kevin Towers said. “He just cooled down after that. There came a time where we were in a pennant race and we had to go with the hot hand. At that time we felt Geoff Blum gave us a chance to win ballgames. That’s why he got the playing time toward the end of the season.

I don’t know about going with the “hot hand,” but Giles was pathetically awful (almost enough so to make me shutup about playing Geoff Blum). In his third straight decline year, he hit .229/.304/.317 in 476 PA. His bb rates and k rates have remained relatively constant. The big difference comes when Marcus puts the ball in play. Here are some numbers:

04
BABiP: .365
Hr/fb: 7.4%

05
BABiP: .337
Hr/fb: 8.0%

06
BABiP: .306
Hr/fb: 6.9

07
BABiP: .275
Hr/fb: 3.3%

He just appears to be putting the ball in play with less force, although there is a good deal of randomness in those numbers. Anyway, I wouldn’t write Giles off, at least not completely. I think he could become the 06 version again, or somewhere around there. But I don’t disagree with the move. $4mill isn’t something you can just throw away (in this payroll range), and when you have a young replacement like Matt Antonelli, letting Giles go is all the more enticing. Good luck to Marcus regaining some of his old form. He’s fun to watch when he’s going good.

****

As I talked about the other week, I expect the Padres to make some moves in the pen (and some that may go under the radar). Here’s part of what I said:

But as we know, relievers are a tough breed to peg and a couple of these guys are bound to have down years. It wouldn’t surprise me to see the Padres make a move or two (perhaps under the radar) to pick up a few more arms in the pen. When you’re so damn good at it, why stop? And they very well might come in handy.

Well, picking up Mauro Zarate from the Marlins off waivers certainly isn’t a big move, but it may pay dividends. In 06, as a 23 year old, Zarate had a solid year in A ball. Here’s his line:

79 inn, 80 k’s, 30 bb’s, 7 hr

This year, he moved from high A ball to AAA and, as you may have guessed, performed quite well:

A+: 26 inn, 20 k’s, 11 bb’s, 1 hr
AA: 25.7 inn, 32 k’s, 9 bb’s, 2 hr’s
AAA: 34 inn, 23 k’s, 12 bb’s, 2 hr

His BABiP was relatively low at each stop and overall checked in at .255 in 2007. That’s something to keep in mind, but still those peripherals are pretty impressive. Here’s his complete minor league line:

278 inn, 247 k’s, 89 bb’s, 31 hr, 3.66 era

Baseball America doesn’t mention him in the 07 handbook (at least it doesn’t on the Marlins minor league depth chart), though, that was of course published before this year. So I’m guessing the scouting perspective isn’t great, but some guys just get overlooked early on.

Anyway, as we’ve talked about, there’s a good chance this move doesn’t work out. It’s a risk in that the probability of Zarate succeeding is relatively low. However, it’s a risk that doesn’t cost that much if it fails. Making 3-4 of these type of moves a year can net you one good reliever and cost you relatively little. It’s a marginal improvement that makes you just that much better in the long run — maybe enough so to avoid those pesky one game playoffs.

Page 9 of 9« First...«56789