NFL Talk Forums  

Go Back   NFL Talk Forums > NFL Forum > NFL Talk

NFL Talk Discuss anything related to the NFL

Reply
 
LinkBack Thread Tools Display Modes
Old 08-22-2017, 18:44   #1
Rookie
 
Join Date: Aug 2017
Location: Stanford
Posts: 33
jackky is on a distinguished road
On the Consistency of ERA

We realize that ERA isnt an ideal indicator of the pitchers talent level. It depends a great deal on the defense behind the pitcher in question. It depends a lot on luck in getting balls in play to fall in which the fielders are. It depends a great deal on luck to get fly balls to land in front of the fence. This will depend a great deal on luck in sequencing getting hits and walks sometimes where it doesnt hurt too much.
Thats Adam Thielen Jersey why we've DIPS. Stats like FIP, xFIP, SIERA, my recent SERA, and Jonathan Judgeseven more recentcFIP all attempt to more accurately measure a pitchers talent by stripping those things out. But what if there is a good way tofigure out just how much ERA actually can differ? How likely a pitchers ERA was? What the spread of po sible outcomes is? These ERA estimators don't addre s that i sue. They can let you know what the pitchers ERA should have been with all the luck removed (or at best the things they think the age should have been), however they cant answer any of the questions I simply posed.
But SERA, which simulates ERA instead of using a formula, could be modified a bit to a sist us out. When we, rather than simulating thousands and thousands of innings at once, break the simulation up into 50- or 100- or 200-inning parts, we can find the distribution of outcomes for your pitcher. This is exactly what I believe the real benefit of SERA is. Its a good ERA estimator not really predictive as xFIP but its biggest a set is the ability to tell us about the variability in a pitchers ERA.
With a new script, we can now set Kevin McDermott Jersey the IP to some lower number, iterate that hundreds or many, many times, and find out the distribution of outcomes. Usually the distribution is not Normal, but luckily most of the time the distribution is around exactly the same. It usually appears like this (the dotted lines are the mean):
Of course, multiplication changes in line with the inputs (it was made using league average because the inputs), the innings per simulation, and even the number of iterations, but its usually pretty similar. The 10th percentile is usually between 1.05 and 1.2 standard deviations in the mean, and it is usually skewed slightly right(that is pretty intriguing; I dont know why that's but my gue s is that really bad values are simpler to get than really good values).
For a typical pitcher, the 10th percentile is simply under 3.00, the 25th percentile is just over 3.25, the medianis about 3.63, the meanis about 3.66, the 75th percentile approximately 4.00, and also the 90th percentile is simply under 4.40. These are kind of like what percentile projections for Pecota do estimate exactly what the best- and worst-case scenarios are. I'll later combine this with pSERA to look at each pitchers uncertainty, in addition to plain mean projection, is for the coming year.
The next thing istrying to determine what makes a pitchers ERA more stable or unstable. Intuitively, you (or at least I) would believe that strikeouts and walks both work to decrease volatility simply because they remove batted ball luck. However, that isn't the case:
The StdDev is the standard deviations from the 1000 simulations of 180 innings each I ranusing the given K% and BB% (with batted ball inputs in a position to league average of 44.8 GB%, 34.4 FB%, 20.8 LD%, 9.6 IFFB%, and 9.5 HR/FB%). You can see that as K% increases so that as BB% decreases, the standard deviation gets lower. This means that yes, a higher strikeout rate does work to decrease the variability in ERA, but a higher walk rate doesn't. This might be since the sequencing luck involved withputting more runners on base includes a greater effect thenthe batted ball luck involved with allowing more contact.
Isaac Fruechte Jersey So, then, going to the K-BB% leaderboardsshould give us a great feeling of whose ERA this past year was a very good indicator of the true talent. The larger the K-BB%, the low variability there is for the reason that pitchers ERA.
What about batted balls? Which batted ball types make ERA probably the most unstable? Heres a chart like the one above, utilizing the same numbers of 1000 iterations of 180 IP, showing GB% and FB% and the resulting standard deviation from the ERAs. LD% isnt shown on the graph, because graphing four variables is hard, but its just 100-GB%-FB%; its lower for the top-right and higher for the bottom-left.
The LD% here's sometimes negative and sometimes crazy high, which is obviously unrealistic. But the overalltrend is consistent throughout: the higher the GB% and FB% and in turn, the lower the LD% the more consistent and fewer variable the ERA is. Its the same DaVaris Daniels Jersey thing as strikeouts and walks. Since line drives turn into hits more often, allowing more of them means more runners on base and also the ERA is much more determined by sequencing luck.
But LD% is generally unreliable and doesnt carry over very much year-to-year. GB% and FB% are much more stable, therefore we need to look at those to see that is more important to reduce variability in ERA that will inform us more about pitchers future ability to maintain a consistent ERA.
Another clear trend. This one implies that a greater GB% along with a lower FB% le sens the variability. I had expected that the higher FB% would lead to a more stable ERA, since more fly balls gives IFFB% and HR/FB% an improved chance to normalize and be closer to the league average. However that isnt the case. When you think more about it, it makes sense, too.
A pitcher having a 50% fly ball rate (which is high) would allow about 285 fly balls over 800 batters, a suming average K%, BB%, and HBP%. Using sampling techniques, we are able to determine that one standard deviation of HR/FB% would be roughly 1.8 percentage points. For a pitcher having a 25% FB% (that is pretty low), the conventional deviation is about 2.5 percentage points. Thats an enormous gap in FB%, although not a major difference within the variationof HR/FB rate. (All this is let's a sume that pitchers do not Cedric Thompson Jersey have control over their HR/FB and IFFB rates, which isnt totally true, but is an a sumption that holds good enough so that you can generalize safely.)
Much more important than allowing HR/FB and IFFB rates to stabilize is preventing big hits like home runs and doubles altogether, since they put more runners on and make more variabilityin sequencing. Ground balls do this very well; almost no ground ball ever winds up as something apart from an out or perhaps a single, savethe occasional one of these. Fly balls, however, find themselves in extra-base hits quite often (over 20% of outfield fly balls go for extra bases).
In a nutshell, I believe the most important takeaway is that this: good pitching = more stable ERA. A better pitcher may have a period that's more suggestive of their actual skill, while a worse pitcher who puts more runners on may have a period that may be quite different from what their actual talent level is. More runners on leads to more uncertainty.The very close correlation between K%, BB%, GB%, FB%, and LD% to the ERA variance tells us that there is a very tangible aftereffect of those things. A higher strikeout and ground ball rate and a lower walk, fly ball and line drive rate not only result in a lower ERA, but also to some much more consistent one.
Next week, Ill check out pSERA for the coming year and employ it to find just how much the age of every pitcher should be expected to vary.
jackky is offline   Reply With Quote
Reply

Thread Tools
Display Modes



Copyright © 2005-2013 NFL Talk. All rights reserved.

Search Engine Friendly URLs by vBSEO 3.6.1 ©2011, Crawlability, Inc.