Adam Burke shares MLB stats he uses to handicap baseball
My go-to line when explaining baseball stats and analytics to people is to say that “they quantify what the eyes can see.” Advanced stats simply paint a better picture than the traditional stats and provide more layers of context to help explain how much a player is helping or hurting his team.
You may watch a game and notice that the center fielder isn’t very good defensively. He’s getting a late read on fly balls and doesn’t have the speed to compensate. He’s got a “noodle arm”. Well, the advanced metrics can look at things like “Outs Above Average” or “Defensive Runs Above Average” to show how much worse than “average” fielder that player is.
A lot of traditional metrics fall short. Think about on-base percentage, which is walks plus hits divided by plate appearances ((BB + H) / PA). That stat really came to prominence around the time that the Michael Lewis book Moneyball was released in 2003. Batting average had always been the go-to stat, but the whole goal as a hitter is not to make an out. Walks are a big part of that equation, thus OBP became a better indicator of offensive output than BA.
But, let’s think more about OBP. Should a home run count the same as a single? Should a double count the same as a walk? With OBP, there is no distinction between the ways of reaching base. Enter wOBA, or weighted on-base average, which is an OBP variant that assigns run values to the manner in which a hitter reaches base. For example, in 2022, a single was worth .884 runs, while a home run was worth 2.072 runs in the formula for wOBA. A walk was worth .689 runs, but a double was worth 1.261. Why? Because doubles increase the chances of scoring a run or driving in a run more than a walk does.
The way I analyze baseball is to utilize the metrics and the statistics to try and find betting opportunities. I don’t want readers or listeners to be overwhelmed by the numbers. It’s my job to explain what they mean, why they are important and how we can use them in our handicapping to find smart bets to make.
I know these stats and abbreviations can be daunting. My only request is that you keep an open mind throughout the season.
The following are some of the stats that I use in my team previews and will use throughout my daily article and my on-air appearances over the course of the season:
Alternate Team Standings Metrics
BaseRuns: In my season previews, I talk about the BaseRuns record. BaseRuns is a “context-neutral environment” that takes all of the individual outcomes, puts them together and spits out expected runs for and against.
Think about these six events in an inning: HR, 1B, 1B, K, K, K. Depending on the order, a team might score three runs (1B, 1B, K, K, HR, K), two runs (1B, HR, K, 1B, K, K), one run (HR, 1B, 1B, K, K, K). The same six outcomes, just in different orders. BaseRuns removes “sequencing”, which is largely random. This also eliminates the randomness of hitting or defending with runners in scoring position. A big disparity between actual record and BaseRuns record might show that a team got lucky or won a lot of close games.
Pythagorean Win-Loss: Pythagorean Win-Loss is a standings metric based on run differential. A team’s runs scored and runs against are put into a formula and an expected record is produced. This concept is used across all sports and is sometimes just labeled “Expected W-L”.
Team and Player Stats
FanGraphs Wins Above Replacement Player (fWAR): Wins Above Replacement Player (WAR) is one of the most comprehensive individual player statistics available. It shows how much value a player has provided relative to his peers. The “f” stands for FanGraphs, which is where I pull my player WAR data from, since I think they do the best job of calculating it.
A player with a fWAR under 1.0 is a replacement-level player, or not a very good one. A player with a 1-2 fWAR is a decent role player or a platoon type of guy. A player with a 2-3 fWAR is a solid player that any team would be happy to have. The scale goes up from there and separates the bad players from the good players and the good players from the great players.
Weighted On-Base Average (wOBA): wOBA is one of my favorite stats because it does add more context than on-base percentage (OBP). Much like OBP, a high wOBA is good and a low wOBA is bad for hitters. On the flip side, a low wOBA is good and a high wOBA is bad for pitchers. Once again, it just adds another layer of context where the outcomes of reaching base are weighted differently based on their importance.
Weighted Runs Created Plus (wRC+): Using wOBA, we can get to a stat called wRC+. With this stat, a player that has a 100 wRC+ is exactly league average. A player with a 120 wRC+ is 20% more productive than a league average player. A player with an 80 wRC+ is 20% less productive than a league average player. Anything over 100 is good, anything under 100 is bad.
The “Plus” means that the stat is park-adjusted and league-adjusted for the run environment. “Park-adjusted” means that there are factors within the formula that account for the way that Coors Field is a better hitter’s park than Miller Park or that Fenway Park is a better hitter’s park than Tropicana Field.
“League-adjusted” basically means grading on a curve. If a lot of runs are being scored, batters have to be that much productive to be above average. If it is a low run environment, then great hitters will stand out that much more.
K% and BB%: I don’t like using raw strikeout and walk numbers or strikeouts per nine innings (K/9) or walks per nine innings (BB/9). I prefer to use K% and BB%, which is simply the percentage of plate appearances that end in a strikeout (K%) or a walk (BB%). This can be used for both pitchers and hitters.
Last season, the league average K% was 22.4% and the league average BB% was 8.2%. This is the first time since 1998-99 that the league K% has dropped in consecutive seasons.
Fielding Independent Pitching (FIP): ERA is subject to a lot of things that a pitcher can’t necessarily control, like batted ball luck and sequencing (the timing of hits, are there runners on base, etc.). FIP takes fielders and balls in play out of the equation by producing a pitcher metric using strikeouts, walks, hit by pitches and home runs, which are believed to be things a pitcher can control.
There is a follow-up metric called xFIP, which stands for “Expected Fielding Independent Pitching”, which is calculated by assuming a pitcher has a league average home run rate per fly ball rate (HR/FB%). This is an indicator used by a lot of bettors that move lines. The idea is that a pitcher with a low ERA and a high xFIP is overperforming and is in line for “Negative Regression” and a pitcher with a high ERA and a low xFIP is underperforming and is in line for “Positive Regression”.
I do a lot of regression analysis handicapping, looking for stats and metrics that focus on pitchers that are seemingly getting lucky or unlucky.
Batting Average on Balls In Play (BABIP): This is a really strong indicator of luck. Think of this as batting average minus strikeouts and home runs. A strikeout is not a ball in play and a home run cannot be fielded, thus it is not a ball in play. The league average BABIP last season was .290. Extremes one way or the other are likely to “regress to the mean”.
Left On Base Percentage (LOB%): This is different from LOB in the box score. This is calculated using a pitcher’s actual hits, walks and runs allowed and the percentage of runners that they have stranded. Think about ERA and how subjective it can be to something like this. The timing of hits (sequencing) matters a lot. LOB% can be a really good stat to use for positive or negative regression.
League average was 72.6% last season. Like BABIP, extremes one way or the other are likely to regress. High-strikeout pitchers typically carry higher LOB% marks because they are going to strand more runners than guys that allow more balls in play.
GB%/FB%/LD%: These are batted ball types, with ground ball percentage (GB%), fly ball percentage (FB%) and line drive percentage (LD%). These are the percentages of balls in play and the distribution of each type. Guys with a high GB% are likely to give up more hits, but fewer home runs. Guys with a high FB% will give up more home runs, but fewer hits, because more ground balls go for hits than fly balls. Line drives are bad for pitchers to give up because they are harder to field.
The league average GB% last season was 42.9%.
HR/FB%: As mentioned with xFIP, HR/FB% is home run per fly ball percentage - How often does a fly ball become a home run. The league average last season was 11.4%. Anything on the high end is likely to come down and anything on the low end is likely to go up.
There are major exceptions. Pitchers with a high GB% could have a higher HR/FB% because they don’t have the sample size of fly balls needed to lower the rate. Similarly, pitchers with a high FB% may give up a lot of homers, but have a lower HR/FB% because they have a larger sample size of fly balls.
You can see most or all of these stats at places like FanGraphs (for a library with more detailed explanations - https://library.fangraphs.com/) or Baseball-Reference.
In recent years, more and more data has been posted for public consumption. Statcast data falls under that category. Housed at BaseballSavant.com, you can see a lot of really detailed statistics for both pitchers and hitters.
I won’t hit on them all, but here are several that I use:
Average Exit Velocity (EV or exit velo): This is a measure in miles per hour of how hard the average batted ball is hit. League average last season was 88.6 mph. Pitchers that are 90 mph or higher give up a lot of hard contact, which is harder to field and often much more damaging. Pitchers that are 87 mph or lower tend to give up softer contact, which is easier to field and less damaging.
Hard Hit%: This is one of my favorite indicators for pitchers. Hard Hit% is the percentage of batted balls hit at least 95 mph. With each mile per hour increase in exit velocity, a batted ball has a higher batting average and a higher slugging percentage.
For example, batted balls hit at least 95 mph last season led to a batting average of .488 with a SLG of .954. Batted balls hit at least 94 mph led to a batting average of .472 and a SLG of .913. Batted balls hit at least 93 mph led to a batting average of .457 and a SLG of .875.
As you can see, with each mph you go down, the numbers start to get better for pitchers. On the flip side, the higher you go, the worse the numbers get. This is a good indicator of pitchers that are commanding their pitches well. The less hard contact, the better a pitcher’s chances of limiting hits and runs.
Of course, you can also look at this for hitters or teams and see teams that make more hard contact than others. The top teams in Hard Hit% last season? Blue Jays, Braves, Dodgers, Yankees, Twins/Phillies. Some pretty good offenses there. There is a high correlation between contact quality and success.
The lowest teams in Hard Hit% last season? Guardians, Athletics, Reds, Nationals, Angels. Some bad teams in that bunch with one outlier.
Barrel (Barrel%): A “barrel” is a batted ball hit at least 95 mph with an optimized range of launch angle. All you need to know is that a barreled ball has an expected batting average of at least .500 and an expected slugging percentage 1.500, so think doubles and home runs.
There were 9,331 barreled balls last season, leading to a .728 BA and a 2.410 SLG. Pitchers that give up a lot of barrels are not in good shape.
Barrel% is the percentage of batted balls that are “barreled”. Something around or above 9% is pretty concerning here.
Spin Rate: With starting pitchers, I’m always looking for decreases or increases in spin rate. A decrease in spin rate can be a good indicator of injury. When foreign substances were banned last season, we saw a lot of decreases in spin rate because pitchers didn’t have sticky substances to stay on the ball later and create more friction and spin.
Spin rate matters because it affects the movement of a pitch. A fastball with a high spin rate will appear to “rise” because it isn’t as affected by gravity on the way to the plate. Breaking balls with higher spin rates will move tighter and break later. There is a high correlation between pitcher spin rate and hitter success in terms of things like batting average and slugging percentage.
Decreases in spin rate will also affect control and command because pitchers are used to throwing in a certain spot. The ball won’t move as much and won’t do what the pitcher wants. It’s why Coors Field is such a hard place to pitch. The thin air and elevation don’t produce as much friction on the ball, so pitches move less, thus making them easier to hit.
Again, all I ask is that you keep an open mind and take the time to try to incorporate some of these metrics or at least read to understand what they mean and why they are important.