Return to Baseball Leftovers

"Buy low" pitchers

Moderator: Baseball Moderators

Postby biju » Thu Apr 27, 2006 12:36 pm

Here's a couple responses to the recent posts:

- It's been a few days since I've run the numbers again so I wouldn't want to rely on the stale data I have. Unfortunately I don't have the time to do this before I go to work (I'm already running a little late) so it's going to have to wait until tonight.

- Please remember that the "buy low" suggestion isn't a "buy low at any cost". I'm just saying that the guy is going to produce better than his current line indicates. So for example I'm not saying cash in your "sell high" Greg Maddux for Ted Lilly. That's a quick way to lose your league. Instead, I would offer up someone you were going to toss aside anyway or if he's on your wire and you have a hole to fill on your bench go ahead and pick him up.

- Jon Garland, while not my favorite pitcher, is getting very unlucky this far into the year and his WHIP and ERA are reflecting this. Again, it's not a buy low at any cost but it would be hard to let what would have seemed like a no-brainer trade in the beginning of the season pass you by at this point.


sockeye wrote:
MissingOakley wrote:Before I think about making trade offers for these guys... your work is based on the idea that pitchers have little control over whether balls put into play become hits or outs. I've heard this talked about before. I've just never gotten fully comfortable with it. I feel like when a guy like Glavine is throwing well, righties are taking weak swings and hitting weak grounders off the end of the bat... but when a pitcher is struggling everything is line drives and he's lucky if they're caught.

So is your claim that the pitchers on your list will start throwing better pitches, which are more likely to induce weak swings... or that if these pitchers keep throwing exactly the same their WHIP and ERA will probably improve?

And most important, have people been successful targeting buy-low fantasy pitchers with this metric in the past?


I'm a little confused by this post, but i THINK it is the same thing I am wondering.....do some pitchers throw in a way that consistently induces weak swings, and therefore the 30% rate is not as good of a benchmark? Intuitively, this would seem to HAVE to be the case for any pitcher who consistently has below-average ERAs, yet is not a strikeout pitcher - ie, he throws a lot of batted balls that, for whatever reason, do not translate into hits.

I would very interested in testing this. I think a nice little study would be to identify the group pitchers have had significantly (statistically, that is) lower hit rates than the league average for, say, 3 years in a row. Maybe going back to 1990 or something to keep it simple. It will probably be a small group, and probably will exclude some high-K pitchers. Greg Maddux in his dominant years would likely to be on it.

If a pitcher cheats the odds consistently enough (and 1 full season is compelling evidence, so 3 would be great), then there is somethign to the idea that a guy with a low hit % may NOT be at the tail end of the normal curve because of chance (and thus likely to revert to the norm), but instead because he has figured somethign out.

And of course, the point of all this would be to get a better idea of whether to grab the guy whose rates have been low for awhile, vs. leaving him on the ww.


I would be willing to bet that the numbers for Hit Rate% will still be around 30% for him. Without looking at the numbers I can only guess he doesn't deviate heavily from the average of 30%. Keep in mind that ERA doesn't correlate with Hit Rate%, otherwise everyone with similar strikeout rates would roughly have the same ERA and that's not the case. Besides, Hit Rate% doesn't (by nature) include walks so a pitcher could walk the bases loaded (still effectively zero Hit Rate%) then get one hit (two runs in), a groundout to 2nd (runner advances to 3rd), a sac fly (third run in), and another groundout. That's a 33.3% Hit Rate% which is high but not too terrible but would basically kill ERA and WHIP.

See what I'm getting at?
biju
Major League Manager
Major League Manager

User avatar

Posts: 1073
(Past Year: -4)
Joined: 6 Mar 2004
Home Cafe: Football

Postby Jaybo1015 » Thu Apr 27, 2006 1:56 pm

how about a top 10 of buy low hitters? :-D
Jaybo1015
Softball Supervisor
Softball Supervisor

User avatar

Posts: 36
Joined: 1 Oct 2005
Home Cafe: Hockey

Postby dc928 » Thu Apr 27, 2006 2:05 pm

;-D
dc928
Major League Manager
Major League Manager

User avatar

Posts: 1244
(Past Year: -3)
Joined: 22 Mar 2006
Home Cafe: Baseball

Postby biju » Thu Apr 27, 2006 2:51 pm

Ok, the boss is out today so I re-ran the numbers here at work. :-D

It's funny because I really thought the "Sell High" list would cause a greater interest but I've since realized that people aren't willing to give up on guys who are winning for them. Unfortunately I think they'll soon see the error in their ways...

If you didn't see it, here's the sell high list:

http://fantasybaseballcafe.com/forums/v ... p?t=194479

Now for some more answers...

tuff_gong wrote:Should we add Jake Peavy to this list?


Peavy's numbers look really good to me all the way around, but his Hit Rate% isn't exaggerated much. He's sitting at 32% which is high but not overwhelmingly so. His HR/Game and K/BB rates are good and he's striking out .77 hitters/inning which is lower than the past two years but not too much. If that corrects expect his Hit Rate to fall right in line.

This leads me to believe three things about Peavy without looking at this seasons actual games:

1. Walks and Hits against him come in bunches followed by stretches of innings where he blanks hitters. Maybe he's getting a little unlucky during inopportune times, or maybe he's throwing worse when there's RISP.

2. He's had a couple of really bad outings to raise his ERA and WHIP up to those levels.

3. In the games he's pitched well in he's not getting proper run support.

PlayingWithFire wrote:just curious, what's Elizardo Ramirez's numbers? He hardly walks a soul and doesn't strike out anybody(around 5k/9ip)


It's tough to get a gauge on pitchers that have little data against them. It would be better to to let him have two more starts or so to get a good feel for him. As for right now with the limited data he's getting lucky more than he should statistically.

MissingOakley wrote:Anyone buying low on Matt Cain?


Matt Cain's numbers look great across the board and he's right in line with 31% Hit Rate (which actually should even drop a little bit). He certainly could use fewer walks and he probably will increase the Ks since he's generally been a 1+ K/INN guy and he's at .74 K/INN instead.

Looks again like he's got the same problems Jake Peavy has with the exception of him handing out more free passes (on a BB/INN ratio) to first base.

kingctb27 wrote:Doug Davis should be atop this list for sure. ;-D


Doug is certainly getting hit more than he should (35% Hit Rate) but part of the problems he's having in the ERA and WHIP cats are due to the tremendous amount of walks he's dishing out. 21 walks in 24.7 INN with 20 Ks isn't going to win you any awards. A good chunk of those came in his last outing and severely raised all of the mentioned categories, but the walks have been fairly steady anyway. Those walks are going to come back to reality however which will reduce the ERA and WHIP but I might wait on seeing improvement *before* trying to acquire him. I'm not trying to start up rumors or anything but he's had back to back 200+ inning season and I'm wondering if he might be an injury waiting to happen (which could even explain the increased walk rate). Again, I don't have anything to back that statement up but that's why I would wait to make sure.

Jaybo1015 wrote:how about a top 10 of buy low hitters? :-D


This is something I've been thinking about a little and while at first I thought I wouldn't be able to pull it off I just now had a slight revelation. I probably won't be able to get to this until the weekend however.
biju
Major League Manager
Major League Manager

User avatar

Posts: 1073
(Past Year: -4)
Joined: 6 Mar 2004
Home Cafe: Football

Postby The Loveable Losers » Thu Apr 27, 2006 3:17 pm

Nice work!

I'd through in a couple extra numbers for each pitcher. Here's a run-down of my formula for predicting whip in an article I've written. I'd use the formula at the bottom to predict the number of hits the pitcher should have given up since you can vary the hit rate with that based on whether they're a fly ball or ground ball pitcher. If you just want the formula it's as follows:

Projected Hits = (BABIP / 1 - BABIP) * (3 * IP - K) + HR

You can take that formula to project the hits a pitcher should have given up and then add the bb's to that to compute their projected whip.

The other formula is Defense Independant Component ERA and can be used to project ERA independant of luck. Here's that one:
DICE = 3.00 + ((13 * HR + 3 * BB + 3 * HBP - 2 * K) / IP)

I find that using those two together is a good way of quantifying how lucky (or unlucky) a given pitcher has been.
Hope springs eternal
grows like ivy on the vine
then dies in the fall

--Haiku for a Cubs Fan, The Loveable Losers
The Loveable Losers
Hall of Fame Hero
Hall of Fame Hero

User avatar
CafeholicFantasy ExpertCafe WriterCafe RankerMock(ing) DrafterWeb Supporter
Posts: 7525
(Past Year: 235)
Joined: 30 Mar 2005
Home Cafe: Baseball
Location: Cubs Win!!! Cubs Win!!!

Postby biju » Thu Apr 27, 2006 3:28 pm

The Loveable Losers wrote:Nice work!

I'd through in a couple extra numbers for each pitcher. Here's a run-down of my formula for predicting whip in an article I've written. I'd use the formula at the bottom to predict the number of hits the pitcher should have given up since you can vary the hit rate with that based on whether they're a fly ball or ground ball pitcher. If you just want the formula it's as follows:

Projected Hits = (BABIP / 1 - BABIP) * (3 * IP - K) + HR

You can take that formula to project the hits a pitcher should have given up and then add the bb's to that to compute their projected whip.

The other formula is Defense Independant Component ERA and can be used to project ERA independant of luck. Here's that one:
DICE = 3.00 + ((13 * HR + 3 * BB + 3 * HBP - 2 * K) / IP)

I find that using those two together is a good way of quantifying how lucky (or unlucky) a given pitcher has been.


Awesome, I love new numbers to crunch. Give me a little time to digest the ideas and run them through the spreadsheet.

thanks!
biju
Major League Manager
Major League Manager

User avatar

Posts: 1073
(Past Year: -4)
Joined: 6 Mar 2004
Home Cafe: Football

Postby The Loveable Losers » Thu Apr 27, 2006 3:33 pm

If you just want to use one constant for all pitchers I've found 0.4 as a replacement for the (BABIP / 1 - BABIP) to work out pretty well as a league average. I think it works out to a .286 BABIP or so. That's in the article but I figured I'd toss that out there as 0.3965 ended up being the best fit to the data (all pitchers from 2003-2005) and I just rounded it to 0.4 so it would be easy to remember and type into a calculator.
Hope springs eternal
grows like ivy on the vine
then dies in the fall

--Haiku for a Cubs Fan, The Loveable Losers
The Loveable Losers
Hall of Fame Hero
Hall of Fame Hero

User avatar
CafeholicFantasy ExpertCafe WriterCafe RankerMock(ing) DrafterWeb Supporter
Posts: 7525
(Past Year: 235)
Joined: 30 Mar 2005
Home Cafe: Baseball
Location: Cubs Win!!! Cubs Win!!!

Postby The Loveable Losers » Thu Apr 27, 2006 4:31 pm

garf112 wrote:
hoya1982 wrote:Edit: Here is something interesting, seems like almost all of Oakland's pitchers were "lucky" last year, this probably reflects the fact that Billy Beane focused on acquiring undervalued defensive players, and thus more balls put in play were fielded by Oakland's defense than is normal.


I think that the fact that foul territory at McAfee Stadium is so huge accounts for almost two extra outs per game for pitchers there. That would mean that there are two extra balls in play that have no chance of becoming hits. That might account for some of the luckiness of Oakland's pitchers, and defense might have some effect on these numbers as well.


While I've never seen it quantified I believe it would be very worthwhile if someone was able to have a 'park factor' and 'fielder factor' that you could use to adjust the coeffecient in my formula. Have them all as numbers fairly close to 1 (especially for the fielders on the team) and multiply them by 3/7 (the actual coefffecient for a .300 BABIP) to get the coeffecient to use for the projections. You could also give pitchers themself a factor directly based off of GB% (the lower the better for them).

For example, say that you're looking at McAfee stadium...let's give it a factor of 0.95. That would take the original coefficient of 0.42857 and adjust it down to 0.41714. Our goal would be to come up with a set of factors for the parks and for each player (obviously each player would have a smaller effect since we'll be factoring in every player on the defense) as well as for ground ball pitchers vs fly ball pitchers (which would also determine how much weight to give to the infielders versus the outfielders).

A factor of 1 would indicate neutrality, above 1 would indicate a player with bad range (errors don't matter here) or a park that's conducive to hits in play (home runs don't matter here as we're only looking at balls in play). Also a factor of exactly 1 would indicate an extreme ground ball pitcher. A factor below 1 would indicate a player with a good range or a park (like McAfee) that's detrimintal to hits in play. Also I think an extreme fly ball pitcher would get a factor of about 0.92 to simulate the effect of a fly ball pitcher giving up less hits in play.

Obviously this would be a massive undertaking that would require ranking all players in baseball (or each defense in baseball from an if/of standpoint at the very least) for their range as well as a study of each park's BABIP average over the course of several years.

For my own purposes 0.4 works just fine for the coefficient but someone with a LOT more time and resources could create one heck of a tool here...to the point where you could actually predict the effect signing a big-name defensive player (like Orlando Hudson) would have on a pitcher's whip.
Hope springs eternal
grows like ivy on the vine
then dies in the fall

--Haiku for a Cubs Fan, The Loveable Losers
The Loveable Losers
Hall of Fame Hero
Hall of Fame Hero

User avatar
CafeholicFantasy ExpertCafe WriterCafe RankerMock(ing) DrafterWeb Supporter
Posts: 7525
(Past Year: 235)
Joined: 30 Mar 2005
Home Cafe: Baseball
Location: Cubs Win!!! Cubs Win!!!

Postby acsguitar » Thu Apr 27, 2006 4:44 pm

Cordero :~(
I'm too lazy to make a sig at the moment
acsguitar
Hall of Fame Hero
Hall of Fame Hero

User avatar
Graphics Expert
Posts: 26605
(Past Year: -117)
Joined: 7 Apr 2004
Home Cafe: Baseball
Location: Back in General Talk WOOO!!!

PreviousNext

Return to Baseball Leftovers

Who is online

Users browsing this forum: No registered users and 0 guests

cron