Clay Davenport Projects Mariners to Win 68 Games

Tied for 2nd worst with the O's and Twins, above the Astros. Ouch.

Clay Davenport is a co-founder of BP for those who don't know.

4 months ago valencia 48 comments 0 recs |

Comments

Those projections seem completely unbelievable.

He expects the Athletics to win six more games than last year, but the Angels will have the exact same record? Give me a break.

by ThirteenOfTwo on Jan 22, 2012 10:55 PM PST reply actions

And the Rangers are supposed to win 102 games.

I don’t think I’ve ever seen a projection system have any team at more than 100 wins before. I shall remain skeptical until I see his methodology, and perhaps even then.

by WhyGodWhy on Jan 23, 2012 12:17 AM PST up reply actions

This guys hitting projections seem overly pessimistic... even for the Mariners

Ackley gets a .248/.325/.354 triple slash with only 7 HR. Even Bill James’ projected sophomore slump didn’t go that far. Clay predicts Montero to be the teams best hitter, but is hardly generous with a .263/.317/.425 line and 18 HR. Smoak’s .250/.337/.401 with 18 HR is favorable compared to other projections (on fangraphs) but that’s it. I took a quick look at just one other team, the Rangers, and it does seem that this guy is generally conservative with the batting numbers but I think he takes that to an unfair extreme with the Mariners.

In short, the Mariners are relying on two very young and still unproven bats in Ackley and Montero to anchor the lineup. Young developing players are hard to project. If Clay Davenport’s system (assuming he has one and isn’t just pulling numbers from thin air) devalues young players then of course the Mariners are going to come off poorly.

I don’t know who Clay Davenport is but I’m not putting much stock in this projection. That is, I don’t put much stock in ANY projection, but I’m especially not putting much stock in this one. The word stock is starting to look weird to me.

by pixburgher on Jan 23, 2012 4:23 AM PST reply actions

Well I decided to scroll down a bit further and take a look at his pitcher numbers.

And now I’ve entirely removed every last bit of my stock from this now entirely stockless projection system.

He has Felix Hernandez inexplicably reverting to 2008 Felix, which is not exactly bad, but is not at all the same Felix we’ve seen for the past three years. I don’t know what the reason is for this predicted decline, but at least we can take solace in the fact that Iwakuma will pitch only slightly worse than Felix. Wait, what? Okay it’s an exaggeration but Iwakuma’s numbers do seem a bit optimistic. Not that I disagree with them. It’s just a bit surprising. 6.4 K/9 and only 2.3 BB/9 are pretty solid numbers to predict for a pitcher moving to the major leagues. So we could call Felix’ projection overly cautious, but then that doesn’t exactly mesh with the Iwakuma porjection.

So how ‘bout those Rangers? The first thing you notice is that they have the best pitcher in all of Major League Baseball. Yes, that’s right. In his first year as an MLB starter, Yu Darvish is going to come in and run away with the Cy Young. I had to check. Justin Verlander, Jered Weaver, Jon Lester—hell even Lee and Halladay—Yu Darvish will outpitch them all. That’s beyond optimism. Again, this is not beyond the realm of possibilities, but to project such success from a rookie? And it makes those conservative batting numbers even more confusing. As for the rest of the Rangers staff, marked improvements projected for Lewis and Feliz. Holland declines a bit.

These pitching numbers are all over the place. Maybe Clay Davenport has a system. Maybe he’s smarter than all of us. But his projections just seem inconsistent and wild.

So why have you subjected us to this madman’s ravings, Valencia? Are you Clay Davenport??

by pixburgher on Jan 23, 2012 5:23 AM PST up reply actions 2 recs

Aaaaand I just realized you told us who Clay is in your post.

Boy is my face red!

…but I still don’t trust his projections. (no stock)

by pixburgher on Jan 23, 2012 5:45 AM PST up reply actions

Figgins will have a higher WORP than Ichiro?

by SgtSasquatch on Jan 23, 2012 6:49 AM PST reply actions

WARP and Eqa instead of WAR and wOBA, really?

Anyway, 68 games is like 21 WAR, right? Let’s do a quick (pessimistic) calculation…
Felix: 5
Vargas: 2
Iwakuma: 1
Noesi: 1
Millwood/Hultzen/whatever: 1
Bullpen: 1.5

Ichiro: 1
Gutierrez: 1
LF: 1
Smoak: 1
Ackley: 2
Ryan: 1.5
3B: 0.5
Montero: 2
C: 1
Other bench: 0.5

Even doing what I think is a pretty pessimistic (like 30th percentile) projection puts us at 69 wins. Depending on if/when Hultzen/Paxton come up, our pitching could be much better, getting any sort of performance out of 3B could make us a lot better, having Montero do some catching duty could make us a lot better, etc.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 9:01 AM PST reply actions

So many things wrong with these projections.

by ThundaPC on Jan 23, 2012 9:53 AM PST reply actions

My favorite is Yu Darvish winning Cy Young by a mile (70.7 RAR)

by ThundaPC on Jan 23, 2012 10:16 AM PST up reply actions

WARP uses a really low baseline "replacement level" iirc, like 40 wins, whereas Fangraphs and B-R use a more realistic 47-50.

So it’s not quite like projecting him to be a 7 WAR pitcher, but it’s still pretty ridic. I think it’s a stretch to project him at more than a 5.5 WAR season.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 11:43 AM PST up reply actions

Even still, if you browse the numbers a bit, Darvish' 70.7 RAR is the highest projected total

I mean, I didn’t check every team.. only those that I think of with top pitching talent; but as I mentioned above, Davenport predicts Darvish to pitch better than Justin Verlander, Jered Weaver, CJ Wilson, Cliff Lee, Roy Halladay, Felix Hernandez, CC Sabathia, Jon Lester, Josh Beckett….

by pixburgher on Jan 23, 2012 12:14 PM PST up reply actions

To be fair, Darvish's JP numbers are really really good

And if you have to base projections off that you’ll end up with really high numbers.

The Darvish he’s projecting is a 9 K/9, 2.5 BB/9 ground ball pitcher. Considering what Daniels paid for the guy, I’m sure his projections kept spitting out the same thing.

by valencia on Jan 23, 2012 12:26 PM PST up reply actions

It's honestly not a huge stretch.

I think Clay overrates Darvish, and underrates Sabathia, Felix, Halladay, and Verlander.
But it’s not crazy to project Darvish as a top 10 SP for 2012.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 1:08 PM PST up reply actions

Felix's RAR was 50.2 last year

Davenport’s projecting 54.6.

Sabathia, 63.8/53.9
Halladay 67.1/67.1
Verlander 63.2/66.8

So only Sabathia is projected as worse than last year.

by valencia on Jan 23, 2012 1:22 PM PST up reply actions

You need to take another look at the numbers because the scale is different.

Davenport is predicting Felix to pitch significantly worse than last year. Felix’ last 3 seasons have been rather consistent and consistently great. The numbers that Davenport gives are more in line with Felix’ 2008 season when Felix was a a very good pitcher but had not yet made the leap to greatness.

Last Year: 233 IP, 8.55 K/9, 2.58 BB/9, 8.40 H/9, 0.73 HR/9
Davenport: 215 IP, 8.04 K/9, 2.85 BB/9, 8.54 H/9, 0.80 HR/9

He’s expecting Felix to go fewer innings while striking out fewer, walking more, and being more hittable.

As Lailaihei just mentioned above, we’re dealing with different definitions of “replacement” here, so you can’t compare Fangraphs’ numbers from last season to Davenport’s projections straight up.

In fact ALL of those pitchers you mentioned are projected as worse than last year. Sabathia is just projected as significantly worse.

And this is precisely what makes the Darvish prediction so absurd. It’s not that it’s inconceivable or even necessarily improbable—it’s just that it is in stark contrast to the cautious projections given for established pitchers.

by pixburgher on Jan 23, 2012 2:53 PM PST up reply actions

A stretch?

We’re projecting Felix at like five wins

by seattlebruin on Jan 23, 2012 1:07 PM PST up reply actions

If you take the median projection for Felix it's like 5.5-6 wins.

Of course, Darvish is potentially a true talent 3-8 win pitcher, but we have a much worse idea of how good he actually is. I would think a reasonable 50th percentile projection could be like 5.5 wins. Of course, there’s all the stuff about adjusting to the ball, parks, batters, etc, which makes it hard to pinpoint what we can actually expect him to do in a healthy season, but I think the systems that are projecting him at 4 wins are selling him a bit short.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 1:11 PM PST up reply actions

A reasonable median projection for a rookie pitcher from the Japanese league is Felix Hernandez.

No. Just, fucking no.

by abender20 on Jan 23, 2012 1:17 PM PST up reply actions 5 recs

Anyone using a projection system on a Japanese player's an idiot.

by Graham MacAree on Jan 23, 2012 1:23 PM PST up reply actions

Projection system: slurp slurp slurp slurp

by abender20 on Jan 23, 2012 1:26 PM PST up reply actions

Jack Z and the Seattle FO are idiots?

Because we signed a Japanese pitcher this off-season and I’m pretty sure we used projections…

by valencia on Jan 23, 2012 1:26 PM PST up reply actions

Our international scouting is, after all, legendarily poor

by Graham MacAree on Jan 23, 2012 1:27 PM PST up reply actions

that doesn't mean we didn't use projections

unless you think the Mariners FO doesn’t believe sabermetrics

by valencia on Jan 23, 2012 1:29 PM PST up reply actions

No

I just believe they’re good sabermetricians.

by Graham MacAree on Jan 23, 2012 1:35 PM PST up reply actions 2 recs

this is going nowhere

according to you: anyone using a projection system on a JP player is an idiot.

That means not only ZIPS, THT, and BP, but the every single FO in the MLB.

by valencia on Jan 23, 2012 1:39 PM PST up reply actions

Clearly I was too vague in my original comment

Anyone taking a projection from the J-League seriously except for the express purpose of sabermetric research is calamitously stupid.

by Graham MacAree on Jan 23, 2012 1:43 PM PST up reply actions

I don't think anyone's taking it seriously

except FO guys, because it’s their job. which once again, puts you thinking MLB FOs are stupid.

by valencia on Jan 23, 2012 1:47 PM PST up reply actions

You seem to be confused about to this

Front offices will rely on scouting to sign J-League players. Then again, you seem to think you know far more than you actually do, so this is hardly surprising.

by Graham MacAree on Jan 23, 2012 1:53 PM PST up reply actions

You are misinterpreting Graham's comment quite badly

1) anyone who bases their evaluation of a Japanese player entirely on a projection system converting to MLB stats is wrong
2) the Mariners likely did project Hisashi Iwakuma over to MLB
2A) the Mariners projection system for Hisashi Iwakuma was very likely scouting based rather than a MARCEL or PECOTA type statistical projection

by seattlebruin on Jan 23, 2012 2:08 PM PST up reply actions 2 recs

"Felix Hernandez", no, but Felix's worst year out of the last 3? Maybe.

Last year, Ian Kennedy put up a 5.0 WAR with 222 innings with about 8.0 K/9 and 2.2 BB/9. I think that’s about what he’ll put up, except with better ground balls, and possibly a worse strikeout rate. Possibly a few less innings, but he’s averaged over 200 the last 3 years, going over 230 last year.

If I were to give a 1-WAR spread around where I think Darvish will end up, it would probably be between 4.3 and 5.3 wins.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 1:30 PM PST up reply actions

So we should place more weight on 2009 and 2010?

That goes against the theory of moving average. Felix’s 3 year RAR average is 57.3 btw. No much farther than the “projected” 54.6. If you add moving average it’s probably right in line.

Dice-K’s K/9 his first year was 8.84. I have a hard time seeing Darvish much lower than that. I think his groundball rate and command is overrated though. I don’t think it’s possible to have 2.54 ERA pitching as a Ranger.

by valencia on Jan 23, 2012 1:44 PM PST up reply actions

I don't like RAR at all.

I project Felix will put up something like a 3.04 FIP over 230 innings, which puts him solidly in the 5.5-6 WAR range.
My best Darvish guess is 4.3-5.3 WAR, a tier below Felix, but still an all-star caliber pitcher.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 1:58 PM PST up reply actions

3.04?

by abender20 on Jan 23, 2012 2:00 PM PST up reply actions

Basically a weighted average of his last 3 years of xFIP.

And then adjusted down because for the last 4 years (basically the outer limits of where pitching numbers matter at all), he’s had a lower FIP than xFIP. I probably gave him a bit too much credit for that skill, but then again FIP probably undervalues Felix’s groundballs somewhat as well. I’m comfortable with a 5.5-6 WAR projection for Felix, barring major injury.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 2:04 PM PST up reply actions

Just use FIP.

We have FIP data on Felix where we don’t have to let xFiP regress it arbitrarily. xFIP is great when you don’t have enough data on someone and are happy to take the safe assumption that they are average at giving up dingers on fly balls. Felix has been better than average at that for the last 4 years (thanks Safeco and Guti). Use his real data.

by abender20 on Jan 23, 2012 2:18 PM PST up reply actions

That makes sense.

I guess I was over-complicating it because I figured that the HR rate fluctuated more randomly than the other factors in FIP, so I didn’t want to weight that part of it.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 2:35 PM PST up reply actions

You're missing the point here. Or at least, you're missing my point.

Even though you’re not responding directly to me, and I don’t want to speak for others. But here it is:

The issue is not with the Darvish projection in and of itself. It’s a very optimistic projection anyone must admit. But that’s fine. If all Clay Davenport had written was, “I predict Darvish will be successful and here are some sample numbers of what I think we might see from him,” then that would have been fine. It’s not inconceivable or outlandish. Certainly expectations are high for Darvish, and there are reasons for them to be so. And there has to be some chance that he meets, and perhaps even exceeds, those expectations.

The issue is with the Darvish projection relative to all other pitcher projections. Davenport is not just expecting Darvish to be successful, he’s predicting him to come in and immediately be the best pitcher in all of baseball! The issue is that the Darvish projection is boldly optimistic while the other projections are cautious and conservative. This inconsistency throws all of Davenport’s projections into question.

Darvish might be the best pitcher in all of baseball next year. It’s not impossible. But there’s a difference between preseason predictions, where one has the freedom to be bold, and a projection system where one is expected to provide average-case scenarios. Predictions can be best case or they can be worst case, but projections should sit in the middle. Maybe the Darvish projection isn’t exactly best case, but it sure is close enough that it doesn’t fit with the rest.

by pixburgher on Jan 23, 2012 3:17 PM PST up reply actions

ZIPS has him at 4.5 WAR; OLIVER has him at 6.4 WAR

OLIVER actually thinks he’ll have 10.3 K/9, 2.0 BB/9. ZIPS agrees on the 2.0 BB/9, but disagrees with a 7.8 K/9.

In comparison, Davenport’s Darvish is 9.5 K/9, 2.4 BB/9. So he’s actually worse than OLIVER’s projections.

The one thing everyone agrees on is he’s a groundball pitcher, which should help in Arlington.

by valencia on Jan 23, 2012 1:19 PM PST up reply actions

Just for comparison...

Colby Lewis put up a 9.4 K/9 and a 1.7 uBB+HBP/9 in his 2 seasons in Japan, and a 8.8 K/9 and 3.0 uBB+HBP/9 is first year here.
Darvish over the past 2 years in Japan put up a 10.3 K/9 and a 2.0 uBB+HBP/9, and did it over a lot more innings.

Mariners fan in SF :: @Eric_Dykstra

by lailaihei on Jan 23, 2012 1:52 PM PST reply actions