Showing posts with label Run Expectancy Matrix. Show all posts
Showing posts with label Run Expectancy Matrix. Show all posts

Monday, March 09, 2020

WEEK THREE (second half): The BirdsofStray return to the Nest against Central Michigan

Image result for ospreys return to the nest

JACKSONVILLE, FL - The 'Strays returned to the Nest for a series against Central Michigan and went 0-4 across the weekend. 

Currently, at fourteen games played out of 53 games now on the schedule (game against Bradley added on 3/18 from the original 52 games) they are past the first quarter pole for the season. 

Let's review beginning with the offensive side of the ball:


The optimized lineup, without regard to L-R splits, was expected to produce 4.81 Runs/G. In the four-game series, UNF scores 17 runs in 4 games for a 4.25 RS/G. 

As things turned out, not paying much attention to splits was appropriate as CMU pitched LHP 3.2 of the 32 IP, a little over 10% of the time. Their pitching staff apparently leans to the right as much as our lineup leans to the left. 

After the CMU series, the optimized (best) lineup configuration appears to be the following:




I compared my caveman calculations versus noted statisticians like Cyril Morong (cool Lineup Analysis tool from baseballmusings.com) and Tom Tango (author of The Book) to massage the lineup but the best Rx is shown above at 4.57 R/g using my numbers and 4.315 Max Rx using Morong's.  

http://www.baseballmusings.com/cgi-bin/LineupAnalysis.py


So, we're getting closer to 2019UNF Rx, which was 5.56. 

RA/G is still abysmal at 8.07 RA/G, versus 2019UNF which was 4.49. 

From a team standpoint, aside from the run disparity (giving up 95% more than creating), 

UNF is getting out-hit (H's) by 51%, 
struck-out 37% more,
OPS is .650 for, .837 against. Not going to win many like that. 
and committing 79% more errors. 

And that is with an inconsistent official scorekeeping, absent which, the numbers might be worse. There's more but I don't want to beat an injured horse with statistics. 

It's good-news/bad-news that the top ERAs on the staff and the lowest BA against are held by 3 freshman and a JUCO transfer on the ERA side, plus a returnee who is hurt. 

The best BA against postings are from three freshman and one returnee from 2019:

Whitehead .150
Roca .246 
Madonna .275 
Chappell .297 

The team needs better production from upper classmen on the pitching side of the ledger soon. 

Last year's UNF team allowed runs distributed as follows:



With 113 RA this year, almost at 2019 total RA through March.

March was a bumpy ride for the staff last year, albeit at lower numbers, but except for one 10 RA in April, the beatings stopped before the middle to end of March. 

Let's hope that happens on the defensive side of the ball this year as well. 

Maybe all we need is a little warm weather. 

I just find it somewhat ironic that we are complaining about the comparatively cold weather. We are playing teams that are likely doing a lot of baseball work inside basketball gyms or outside in weather that would make an Alaskan dog-sled team shiver.  

In future posts, I want to try to:

illustrate some of the observations on batters splits better,

** 2019 UNF Batter Splits** (added):


** together with the 2019 UNF Batter Splits above you can use the 2020 splits dashboard to track splits and revere splits among personnel. It can be a different lineup depending on availability (health) and the handedness of the pitcher on the mound ** 

*** 2020 UNF Batter Splits Dashboard **:

I am already not a fan of the RC columns since those are cumulative or counting stats. I like the PA versus pitcher handedness to see if there is enough data, given the prior caveats. 


STILL ON The  Slav's TO-DO LIST:
add some work on pitchers splits, 
and 
investigate the success of this guy, freshman Tony Roca:




#32 LHP just carving through batting orders all cool, calm and collected. Dare I say, like "the left arm of God" Sandy Koufax. 
Tony Roca delivers a pitchImage result for sandy koufax

OK, maybe a bit much, but he is making all the "paisanos" from Jacksonville to Ponte Vedra Beach happy. 

Next up, two against Ohio State during the week and a weekend of three versus Illinois State. I know the Redbirds aren't getting any quality outdoor time up there and neither are the Buckeyes. 

I coached a season of freshman baseball in Illinois during the spring months and it was one of the worst winters I ever endured. 


Other good information regarding Lineup Optimization:
 https://www.beyondtheboxscore.com/2009/3/17/795946/optimizing-your-lineup-by
#1, #4, #2, #5, #3, #6, #7, #8, #9 So, you want your best three hitters to hit in the #1, #4, and #2 spots. Distribute them so OBP is higher in the order and SLG is lower. Then place your fourth and fifth best hitters, with the #5 spot usually seeing the better hitter, unless he's a high-homerun guy. Then place your four remaining hitters in decreasing order of overall hitting ability, with base stealers ahead of singles hitters. Finally, stop talking like the lineup is a make-or-break decision. Of course, enough empirical analysis has revealed serious flaws with this approach. Turns out, the two-hole is actually the most important placement in the lineup. The cleanup and number five spots should be occupied by the two next best overall hitters on the team.
Tom Tango, Mitchel Lichtman, and Andy Dolphin used run expectancy to reach this conclusion, evaluating the potential of each position in the order to generate runs. While their overall analysis far exceeds the scope of this article, you can read their book covering the issue (among other interesting sabermetric topics). It’s called The Book.
Using sabermetrics, the following general guidelines are recommended for batting order.
      https://www.sportsbettingdime.com/guides/strategy/batting-order-sabermetrics/
Qualities 1. Highest OBP on team; Speed is important but ability to get on base matters much more OBP1
2. Highest wRC+ on team; Very high OBP; Best overall batter on the team; Must be consistent and balanced OBP2
3. Good overall batter; High wRC+; Less important than #5 spot; Likely 4th best hitter overall wRC+
4. Most powerful hitter on the team: Must have very high SLG%, HR, ISO SLG,HR,ISO
5. Good overall batter with extra emphasis on power. High wRC+; High SLG%; More important than #3 spot; Like 3rd best hitter overall wRC+,SLG
6. Average/good overall batter with emphasis on baserunning; high OBP and SB OBP,SB
7. Average overall hitter RC
8. Average overall hitter RC
9. Worst hitter on the team; likely the pitcher in NL games RC


Sunday, March 08, 2020

WEEK THREE (first half): The BirdsofStray return to the Nest against Central Michigan

Image result for ospreys return to the nest

JACKSONVILLE, FL - The 'Strays return to the Nest for a series against Central Michigan and we'll continue to test drive some baseball statistical concepts in real-time.

Including Games played through 3/6:

UNF Runs scored: 47 
UNF Runs allowed: 87 
Win Probability .246 (13-39 in 52 games)

RCx based on OBA/SPCT: 44.82 
RAx based on OBA/SPCT: 81.50 
Win Prob. Based on Rx: .275 (14-38 in 52 games)

Offense (based on OPS, K%):

Seguera .930 (25%), 
Presno .837 (32%), 
Matthias .769 (31%), 
Grodi .746 (13%), 
Libuano .712 (21%)

Team Average .659 (23%)

T.May .587 (27%), 
West .582 (36%), 
Marabell .573 (23%), 
T.Clark .573 (12%)

  • Too many strikeouts, not enough pop. 
  • Lineup tilts to the left side.
  • Need RH bat to emerge above the team average.
  • Runs have decreased from ~5.5+ per game last year to ~4 per game this year.


Defense:
Runs allowed have increased from ~4.5 per game last year to ~7 per game this year.

The disparity in RC/RA decreased from +1 per game last year to -3 per game this year.

Errors per game have increased from 1.2 per game last year to 1.77 this year.

Bunt defense is still an issue, which it was last year.

Fielding is close to this situation right now:

Pedro Guerrero was playing third base for the Los Angeles Dodgers in 1984 when he made the comment that earns him a place in my Hall of Fame.  Second baseman Steve Sax was having trouble making his throws.  Other players were diving, screaming, signaling for a fair catch.  At the same time, Guerrero, at third, was making a few plays that weren't exactly soothing to manager Tom Lasorda's stomach.  Lasorda decided it was time for one of his famous motivational meetings and zeroed in on Guerrero: "How can you play third base like that?  You've gotta be thinking about something besides baseball. What is it?"

        "I'm only thinking about two things," Guerrero said.  "First, `I hope they don't hit the ball to me.'"  The players snickered, and even Lasorda had to fight off a laugh.  "Second, `I hope they don't hit the ball to Sax.'" 

 -- Joe Garagiola, "It's Anybody's Ball Game"

The pitching was expected to be about the same as last year and the fielding and hitting somewhat worse due to graduation and inexperience.

The starting pitching is virtually in total disarray, save from the performance of freshman Tony Roca.

Except for Ohme, who is hurt, only freshmen and incoming JUCOs are under the average team ERA of 6.43 and they have pitched 43/112 innings total (38%). This represents four starts out of the first 13 as well. 


The efforts from upperclassmen in this area must pick up before conference play or you’re playing for next year. The first third of the season (18 games) appears as if it will be difficult. After that, the schedule is kinder for 12 games or so. Then conference play begins in earnest.

Hitting:

As far as the hitting statistics, 100 PA are better than 50 PA no matter what statistic you are measuring.  It’s always better to have more data, but the rate at which the data becomes useful varies based on the statistic.

The only statistics that stabilize (become reliable) before 200 PA’s are: https://library.fangraphs.com/principles/sample-size/

K-rate (60PA)
BB-rate (120PA)

The rest take longer than one season and season to season comparisons are problematic at the NCAA level. But you must work with what you have to gain insights and make decisions.

By contrast, stats we depend on for insight like OBP & SPCT can take between 320 – 460 PA to stabilize.

Given those caveats, I did pull down the splits versus LHP and RHP for 2019 and 2010 for the Ospreys.  

Last year we did see a couple of LH hitters (Law and Libuano) hit 
LHP’s better than RHP (a reverse split). 

Conversely, Tanner May and Tanner Murphy appeared to do better against RHP than LHP.




This year, in limited action, Chris Matthias appears more comfortable against RHP as does switch hitter Sequera.

LH hitters Grodi, Spratling-Williams, Presno and West to date are hitting LHP better than RHP, which you do like to see somewhat. 

Somebody has to crush RHPs and it’s generally harder to rely on RH hitters to do that. We’ll see how it goes, remember that by the time the data is reliable, the season is over.

You have to know your personnel.

I would prefer to have purely exit velocity splits, but I haven’t seen any site that collects or posts that and you would have to have Statcast-type capability and you won’t see that here due to a leadership weakness called titlulus novem cultus.

Good luck finding the cure for that!!




Wednesday, February 26, 2020

WEEK TWO (and a half): The BirdsofStray, and TOTO too, leave the Nest for a game against South Carolina


Image result for run expectancy matrix

SABERMETRICS 101: RE24

JACKSONVILLE, FL - The 'Strays leave the Nest for their first road trip to Columbia, SC and I used the opportunity to test drive some baseball statistical concepts in real-time via Twitter. 

Using basic Data Analysis tools (Excel, iPhone calculator), we looked at Lineup Optimization, Run Expectancy & Win Expectancy. 

The college formula for RC (Runs Created), would need to be adjusted from 38 Plate Appearance per game to about 40, using last years UNF data and 2019 NCAA pitching data. 

Also noted that the early inning where SC tried to give the game away -- Top of 2nd with the 3 walks and two errors -- really changed the odds significantly, especially the errors. They count as an AB without a hit attached (negative to offenses RE) although they do add to OBA. There should be an additional adjustment to the offense, for adding most significant resource, additional out(s). 

I suppose you would then have to further complicate the formula to account for double plays and caught stealing, or net them out, as outs added back to the offense or given back to the defense. No thanks!! 

For a back of the envelope type of calculation, this is not too shabby. It's hard enough to concentrate with that blasted Run Expectancy Matrix (RE24) running through my head. 

This was the box score from the game and my usual post-mortem bitching and whining on Twitter. 

from ospreys.com

UNF (2-5) -VS- SOUTH CAROLINA (6-2)






UNF logo
South Carolina logo
Team123456789RHE
UNF040010000550
 South Carolina12101100X6123
W:
Julian Bosnic (1-0)
 
L:
KILFOYL, Darin (0-1)
 
S:
TJ Shook (1)






We came in with a lineup that was capable of scoring 3.932 runs & came out with 5. The early walks & errors contributed to the overage, but the 3.932 is a concern. Configure optimally the expectation would be 5.01 runs. We are currently @ xRA 5.98, actual ~ 8 RA, pitching. Ugh!
Quote Tweet
Charles Slavik
@theslav1959
·
FINAL 6-5: They tried to give it away early and did just enough to hold on. Only 4K's to UNF 11K hitting. 12H to 5H and 3xBH to 1xBH. OBA SPCTPA Rx Wx 0.359.......0.182........38.00.........2.48.......0.048 0.452.......0.645........38.00.........11.07....... 0.952
2
1





BTW: If the xR was 5.01 coming in...& the opponent bumps that up by 1.068....poof! 6.078 Runs, maybe we're still playing. Or we utilize the bullpen different & sneak away with a W. Just thinking out loud. BTW2: This was very predictable for the program, given larger prior issues.





Replying to
Currently on a glide path for 24-33 record, if we get to 57 games played. In near term, playing #19 Hartford, #148 CM, #86 Ohio St. and #61 IL St. could lead to 3-9 which leaves team 5-14 entering B.C. series and FAMU. A sweep there (??) leads to 9-14 entering conference. Ugh!
1
1





I know some were pissed when Baseball America picked UNF for 5th in conference. That's looking rather generous as right now, Warren Nolan has us ranked #286, last in conference. The trend is not our friend right now.
1



These were the real-time game posts, after each inning. I messed up the formatting on Twitter for the first couple of innings. What a Twit?







FINAL 6-5: They tried to give it away early and did just enough to hold on. Only 4K's to UNF 11K hitting. 12H to 5H and 3xBH to 1xBH. OBA SPCTPA Rx Wx 0.359.......0.182........38.00.........2.48.......0.048 0.452.......0.645........38.00.........11.07....... 0.952







After 8, down 6-5, it's Hail Mary time: OBA SPCT PA Rx Wx 0.389........0.200........38.00........ 2.96......... 0.067 0.452........0.645........38.00........ 11.07....... 0.933






After 7, down 6-5. Good timem for a 2-spot: OBA SPCTPA Rx Wx 0.480.......0.300........38.00........ 5.47......... 0.196 0.452.......0.645........38.00........ 11.07...... 0.804





A big AB by Libunao in T7. Looking like another 1-2-3, low pitch count inning gets extended and almost scores a Little League offense run, BB-WP-WP-(almost....)





Though 6, down 6-5 after 11 pitch T6: OBA SPCT PA Rx Wx 0.480........0.300........38.00........ 5.47........ 0.172 0.448........0.704........38.00........ 11.99...... 0.828





After 5, 5-5. A big hold by Kilfoyl there. Creeping back slowly: OBA SPCT PA Rx Wx 0.480.........0.300........38.00........5.47......... 0.167 0.417.........0.773........38.00........ 12.23..... 0.833















After 2: Up 4-3, W&E skew Rx/Wx, less likely to come in bunches at MLB level. It is two different boxscores over SSS of 2 innings. WAYGD? OBA SPCT PA Rx Wx 0.5830.22238 4.92 0.114 0.4000.90038 13.68 0.886
















We came in with a lineup that was capable of scoring 3.932 runs & came out with 5. The early walks & errors contributed to the overage, but the 3.932 is a concern. Configure optimally the expectation would be 5.01 runs. We are currently @ xRA 5.98, actual ~ 8 RA, pitching. Ugh!
Quote Tweet
Charles Slavik
@theslav1959
·
FINAL 6-5: They tried to give it away early and did just enough to hold on. Only 4K's to UNF 11K hitting. 12H to 5H and 3xBH to 1xBH. OBA SPCTPA Rx Wx 0.359.......0.182........38.00.........2.48.......0.048 0.452.......0.645........38.00.........11.07....... 0.952
2
1





BTW: If the xR was 5.01 coming in...& the opponent bumps that up by 1.068....poof! 6.078 Runs, maybe we're still playing. Or we utilize the bullpen different & sneak away with a W. Just thinking out loud. BTW2: This was very predictable for the program, given larger prior issues.





Replying to
Currently on a glide path for 24-33 record, if we get to 57 games played. In near term, playing #19 Hartford, #148 CM, #86 Ohio St. and #61 IL St. could lead to 3-9 which leaves team 5-14 entering B.C. series and FAMU. A sweep there (??) leads to 9-14 entering conference. Ugh!
1
1





I know some were pissed when Baseball America picked UNF for 5th in conference. That's looking rather generous as right now, Warren Nolan has us ranked #286, last in conference. The trend is not our friend right now.
1

Giants Top Minor League Prospects

  • 1. Joey Bart 6-2, 215 C Power arm and a power bat, playing a premium defensive position. Good catch and throw skills.
  • 2. Heliot Ramos 6-2, 185 OF Potential high-ceiling player the Giants have been looking for. Great bat speed, early returns were impressive.
  • 3. Chris Shaw 6-3. 230 1B Lefty power bat, limited defensively to 1B, Matt Adams comp?
  • 4. Tyler Beede 6-4, 215 RHP from Vanderbilt projects as top of the rotation starter when he works out his command/control issues. When he misses, he misses by a bunch.
  • 5. Stephen Duggar 6-1, 170 CF Another toolsy, under-achieving OF in the Gary Brown mold, hoping for better results.
  • 6. Sandro Fabian 6-0, 180 OF Dominican signee from 2014, shows some pop in his bat. Below average arm and lack of speed should push him towards LF.
  • 7. Aramis Garcia 6-2, 220 C from Florida INTL projects as a good bat behind the dish with enough defensive skill to play there long-term
  • 8. Heath Quinn 6-2, 190 OF Strong hitter, makes contact with improving approach at the plate. Returns from hamate bone injury.
  • 9. Garrett Williams 6-1, 205 LHP Former Oklahoma standout, Giants prototype, low-ceiling, high-floor prospect.
  • 10. Shaun Anderson 6-4, 225 RHP Large frame, 3.36 K/BB rate. Can start or relieve
  • 11. Jacob Gonzalez 6-3, 190 3B Good pedigree, impressive bat for HS prospect.
  • 12. Seth Corry 6-2 195 LHP Highly regard HS pick. Was mentioned as possible chip in high profile trades.
  • 13. C.J. Hinojosa 5-10, 175 SS Scrappy IF prospect in the mold of Kelby Tomlinson, just gets it done.
  • 14. Garett Cave 6-4, 200 RHP He misses a lot of bats and at times, the plate. 13 K/9 an 5 B/9. Wild thing.

2019 MLB Draft - Top HS Draft Prospects

  • 1. Bobby Witt, Jr. 6-1,185 SS Colleyville Heritage HS (TX) Oklahoma commit. Outstanding defensive SS who can hit. 6.4 speed in 60 yd. Touched 97 on mound. Son of former major leaguer. Five tool potential.
  • 2. Riley Greene 6-2, 190 OF Haggerty HS (FL) Florida commit.Best HS hitting prospect. LH bat with good eye, plate discipline and developing power.
  • 3. C.J. Abrams 6-2, 180 SS Blessed Trinity HS (GA) High-ceiling athlete. 70 speed with plus arm. Hitting needs to develop as he matures. Alabama commit.
  • 4. Reece Hinds 6-4, 210 SS Niceville HS (FL) Power bat, committed to LSU. Plus arm, solid enough bat to move to 3B down the road. 98MPH arm.
  • 5. Daniel Espino 6-3, 200 RHP Georgia Premier Academy (GA) LSU commit. Touches 98 on FB with wipe out SL.

2019 MLB Draft - Top College Draft Prospects

  • 1. Adley Rutschman C Oregon State Plus defender with great arm. Excellent receiver plus a switch hitter with some pop in the bat.
  • 2. Shea Langliers C Baylor Excelent throw and catch skills with good pop time. Quick bat, uses all fields approach with some pop.
  • 3. Zack Thompson 6-2 LHP Kentucky Missed time with an elbow issue. FB up to 95 with plenty of secondary stuff.
  • 4. Matt Wallner 6-5 OF Southern Miss Run producing bat plus mid to upper 90's FB closer. Power bat from the left side, athletic for size.
  • 5. Nick Lodolo LHP TCU Tall LHP, 95MPH FB and solid breaking stuff.