Sports Statistician Answers Sports Math Questions From Twitter | Tech Support | WIRED
Вставка
- Опубліковано 12 чер 2024
- Statistical analyst and ex-NBA assistant coach Dean Oliver visits WIRED to answer sports math questions from the internet. What’s the most efficient shot to take on a basketball court? Are NFL running backs historically undervalued at the moment? Why are there more no-hitters in baseball lately? And statistically speaking who’s better: LeBron James or Michael Jordan? Answers to these questions and more await on Sports Math Support.
Director: Justin Wolfson
Director of Photography: Constantine Economides
Editor: Matthew Colby
Expert: Dean Oliver
Line Producer: Joseph Buscemi
Associate Producer: Brandon White
Production Manager: D. Eric Martinez
Production Coordinator: Fernando Davila
Casting Producer: Nick Sawyer
Camera Operator: Cloud
Gaffer: Gautam Kadian
Sound Mixer: Rebecca L. ONeil
Production Assistant: Albie Smith
Post Production Supervisor: Alexa Deutsch
Post Production Coordinator: Ian Bryant
Supervising Editor: Doug Larsen
Additional Editor: Paul Tael
Assistant Editor: Billy Ward
Still haven’t subscribed to WIRED on UA-cam? ►► wrd.cm/15fP7B7
Listen to the Get WIRED podcast ►► link.chtbl.com/wired-ytc-desc
Want more WIRED? Get the magazine ►► subscribe.wired.com/subscribe...
Follow WIRED:
Instagram ►► / wired
Twitter ►► / wired
Facebook ►► / wired
Tik Tok ►► / wired
Get more incredible stories on science and tech with our daily newsletter: wrd.cm/DailyYT
Also, check out the free WIRED channel on Roku, Apple TV, Amazon Fire TV, and Android TV.
ABOUT WIRED
WIRED is where tomorrow is realized. Through thought-provoking stories and videos, WIRED explores the future of business, innovation, and culture. - Розваги
It never fails that, the more someone really knows basketball, the more complex and non-commital their answer is re: Jordan vs. LeBron. They're very difficult to compare as players.
Lebron is a beast, but lack charisma ... just look at the movie space jam 2 vs original
That's because he is answering politically, just like Ronaldo vs Messi. Jordan has more hardware both individual as well as team than LeBron and KD put together. Enough for me to pick the best.
This is true of most things in life. The more expertise you have in a topic, the more aware you are of nuances & gray areas, and so the less likely you are to make sweeping claims with black and white thinking.
People are generally suprised to find that LeBron has a higher 2pt FG%, 3pt FG% and FT% than MJ did, and that MJ scored more per game because LeBron always looked for the open teammate first before taking a shot ... at least until the last two years in LA. Also, LeBron wasn't forced out of the league for two years due to gambling. So there's that. MJ was a more dynamic personality and imo more fun to watch. LeBron clearly had the greater total career. MJ had the greater peak.
@@studgerbil9081 so many excuses for the lack of rings and individual trophies LeBitch James has. Jordan did it all and didn't take him 20 years to do it. One thing he does better, I'll admit, no one flops, dives, cries and overacts like LeBron James!
The benefit of switching pitchers more often that was not mentioned is that hitters tend to do much better by the 3rd time they face the same pitcher in the same game, not only the pitcher is getting tired, but the hitters also had 2 previous experiences fresh. By switching to a new pitcher, hitter's previous experience gets nullified.
That's how you used to tell the real aces. Guys who could get through the lineup that 3rd time were the elites
A big reason for no hitters, People are selling out for the homer at the plate and the strikeout as the hitter, this lowers batting averages and raises walks
Another thing is Relievers usually have better velocity and/or movement, so using more of them is often better than letting your starter go longer. However you also run into the problem of either working your relievers too hard and they struggle, or simply using too many pitchers, and only a handful of them are any good. Then it becomes important for starters to throw lots of innings again.
In postseason series, though, this seems to be leading to the problem reoccurring. Instead of a lineup seeing the starter too much, they may be seeing certain relievers too much over the course of a series. And relievers are usually relievers because they have less of an arsenal of pitches than a starter.
This kind of thinking and analytics time and time again lose teams games. As a Dbacks fan, Torey Lovullo has lost us countless games adopting this mentality.
I will forever appreciate Wired for calling it Twitter instead of X.
Move on
@@josemv25dawg if youre calling it X youre really weird
@@zaycomage6888 I don't even use Twitter
Probably filmed before
i'll check back in 1000 years to see if you're still appreciative
i like how the statistician fittingly uses the word "probably" so often
Win probabilities... probably. 😆
probably, approximately, potentially, etc are all the favorites 💀
@@based__381And if you understood anything about statistics you would understand why. But you don’t so you just think it’s funny or out of place or something
@@vinniehuish3987 The only reason i know that those words are so commonly used in stats is because i’m very familiar with the subject. Idk why you would assume im not.
@based__381 well we probably, likely, potentially wont see a rebuttal
I’m very impressed with this man’s knowledge of in depth statistics in several sports. He definitely knows what he’s talking about in Basketball, Baseball, and Football
It’s almost as if it’s his full time job
@@hoptoit8687lol always gonna be a douche like you in the comments huh
He knows a lot about American sports. But saying that the best statistic we have for soccer is number of goals is a terrible answer to the question on ronaldo v messi
@@mrpigcho9466probably because football is not USA's top sport historically
@@mrpigcho9466 Right... which is why he said it is not ideal, but they do not have many other statistics at this time. He said himself that it isn't optimal so Idk what video you watched lol
I was a sports statistician for one of the biggest media companies. Everything this guy just said is accurate, especially the part about how you get into the profession.
didn’t really assume it wasn’t
didn’t really assume it wasn’t
prove it
How exactly did you get into this field, would you recommend this to someone that loves math and watching sports?
Ofc it is. He's like the father of basketball analytics
Would love to see at least one question about hockey. Please bring this guy back. Interesting video, would love a bit of a deeper dive.
Can't believe there was no hockey question...
I would love to hear his take on the best way to measure success for a team / player on the Ice. Is it +/- Score? How does that apply to Goalies?
if knows hockey as much as he knows soccer than his answer would probably disapoint you, no offense to him tho, is amazing being an expert in three diferents sports like him
maybe they just didn't go there but i'd guess he doesn't have much expertise in hockey
@@dizzyharris2658plus minus is widely disregarded in the world of sports analytics. Most people would just as soon toss it out. Bad players on good teams post great +/- all the time with the reverse being true of good players on bad teams.
It’s good seeing tech support videos back. I was worried Wired stopped doing those since we didn’t get any for a few weeks, until we got 1 last week and now this
I would say an underrated WR is average separation gained on a route run. It seems rather consistently that WRs in college with poor separation numbers tend to fail in the NFL
Agreed, that’s a big one. Especially for the college to NFL transition.
I think that this stat gets a bit murky at the NFL level, though. For one, defenses are far more complex - you don’t see a lot of zone/man hybrids in college. It’s hard to know what the goal of the WR even is on each play - find a hole in the zone? Decoy? Purposefully pull help coverage your way? Defenses also play off on purpose at times. This all exists in college, but it’s overall a lot more straightforward in College.
Also, college defenses are, typically, pretty weak. Good WRs often play 1-3 good defenses a year, with the rest being trash. Most NFL defenses are decent quality - not many trash defense games, if any. Which might appear to be an argument for using this stat in the NFL, but it’s really saying that the gulf in quality among college defenses really means that you need a good avg separation there to be considered good enough to compete in the NFL.
- How is Messi better than Ronaldo, statistically speaking?
The sports statistician:
- Goals 🤷
Won't blame him if he's not familiar with football. He's American after all. But for the rest of the questions, he did very well.
This debate was already over last year. It's just Ronaldo fanboys who are still clinging desperately to this question.
I also don’t know much about soccer statistics. I assume they track things like assists and shooting percentages he could have mentioned. But do they also track things like turnovers versus passes, goals versus expected goals, offsides penalties, on/off goal differential, or stops/steals against ball handlers?
xg per 90, figured he would at least know that. or shots per goal ratio
@@acgeewhiz They have plenty of statistics, and yeah they track all of that stuff alongside stuff like key passes and dribbles. American sports are a few years ahead of it in terms of a statistical revolution, partially due to circumstance but largely due to it being a much harder sport in which to quantify impact. Simple put, there's a much wider latitude of what each statistic can mean in soccer compared to, say, baseball where each incident that comprises a single stat has essentially the exact same value.
Also, the game's structure makes it hard to get good statistical samples since there are only ~3 goals per game on average and maybe 10 incidents which decide the match and can accurately be quantified. And most elite players play for the entire game so you can't get reliable on/off data.
Props to him for 99% of this, the football/soccer question about Messi and Ronaldo was very bland and simplistic thought. Especially when stats like xG and xA exist
The question was asking how is Messi better than CR7 then he talked about goals, well, in goals Messi is not better than CR7 lmao
@@sh3ltonyes he is? Messi has a better goals to game ratio than Ronaldo.
@@davies010 ronaldo has more total goals. it's important to mention the positions that ronaldo played. he didn't really play a primary striker role until he got to madrid.
@@huegrection9698 SHUT UP
@@huegrection9698 shall we ignore that Messi play with Suarez for most of him prime then
I’m not a huge sports fan, but I am fascinated with the statistics, analysis, psychology, and other complex factors. It helps with training, recruiting , historical statistics , and other useful outcomes.
No one cares you golorified hr woman
There's better stats than "just" goals to measure a footballer's performance
Like what
I wrote a short paper for a sports analytics class arguing why the three-point shot was the most effecient (not including possibilities of fouls). Glad it was on the right path!
There's a few caveats though, it is more efficient AS LONG AS THE PERCENTAGE IS ABOVE 35%. So there's a few things to consider:
1) A tough, contested three-pointer will not be above 35% regularly, it's probably 20%-25%. So don't chuck forced threes!
2) It depends on how good a shooter you are. Shaq will never hit above 35% even when wide open. So it's not more efficient for him!
This is something that I have always wanted to explore, thank you wired for posting this video!
I would like to see a sports statistician answer whether he believes individual catch-all advanced stats (EPM, LEBRON, RAPTOR, or the more antiquated PER), particularly defensive catch-all stats, are useful to determine how good individual players are.
EPM is the holy grail at the moment brother
@@kingsally2864 What I'm wondering though is an expert's opinion on how good those catch-all stats are at actually capturing the best players.
He mentioned how team advanced statistics (offensive and defensive rating) are better at determining how good a team is than in sports like baseball and football because of the number of possessions. I'm wondering if for the reason that basketball outcomes are relatively more complex than baseball outcomes (5 vs. 5 instead of the "1 vs. 1" dynamic between pitcher and batter), individual statistics like EPM are less predictive of how good a player is than those like WAR in baseball.
@@marcr196 i agree would love to see that, i would watch an entire episode of wired on that
@@kingsally2864 the holy grail right now is darko dpm
There should be more statistics on refs/umps that are more available to the fans.
Wonder how this guy does in his Fantasy League 🤔
Wish he would’ve touched on separation stats when he brought up receivers. Fantasy experts love separation and avg depth of target stats
I could listen to this man all day
I feel like there’s more than just goals to tell how good a player is
Yeah, he knows a lot about sports, but I don’t think soccer is his area of expertise so his answer wasn’t great
Are they a main defender? Or is their main purpose to put the ball in the net? If so, then the player who puts the ball in the net more is the better player
Man there are so SO many metrics in football (soccer) to judge player performance than just "goals". It doesn't even account for the different roles and positions that players play. I love all of his analysis, but this one take is kinda ridiculous.
He's american. Americans do not give a single solitary fugg about football (soccer).
Football is statiscally underexplored though. And most of the current statistics are crap. Like theres no metric to measure a defensive midfield player individually. How he affects the field off ball for instance.
@@gaming4yallexactly. there’s no stat that can determine how well a midfielder dictates tempo or anything else that can only be viewed through the eye test. it’s understandable why they don’t have stats for these types of things
I think he said goals because both of them are attackers. Just like for receivers he said catches and didn't take into account everything the receiver does to get into the position to catch. By listening to some of his other answers it seems that there are in fact basic stats that are "good enough" to determine someone's skill and goals/catches are some of them. If a receiver is good he will get into position to make catches. If a football attacker is good he will find scoring opportunities. Remember Filippo Inzaghi? Not the fastest, not the most technical, but his sense for positioning was through the roof.
@@RunawayYeif that's the case, then he just shouldn't have responded to that question, bc he just doesn't understand the sport
For someone trying to change the stat game in darts, this is a great video to watch! Thanks to Wired & Mr. Oliver for making this!!
For those who are wondering, the best WR metric is Yards per Route Run. It’s pretty much an all encompassing efficiency metric
except that the most important part of the job is actually catching it, ie the Chiefs recently
@@Tassadoniapls don’t remind me😢. Our wr core is so bad this year
IMO the most important part of the video was the little bit he said about triangle bc so many ppl and teams use the “make the defense pick and choose” idea and then the triangle flips that to offenses choosing and flowing. The other side of that r having the smart enough players that r good enough threats on ball to make a triangle function.
Me, a European, listening about baseball and American football: yes, those words probably exist
Most Americans would feel the same when hearing about cricket 😆
Europeans too
Pitcher wins have a correlation with how good they are, but there is also a correlation with how good the rest of the team is.
came here to say this. Take two identical pitchers, the one with a better offense backing him up will have a better W/L record. Simple as that!
It’s why King Felix was one of the best pitchers without a great win count. The Mariners just couldn’t score!
Nolan Ryan had that problem, although he just through so many innings he eventually got there on the counting stats.
@@tjtribble.
I'll argue there was at least a little bit of merit to the Pitcher win before the 80s or 90s, when pitchers would throw complete games with regularity and even then throwing 7 or 8 innings there's a significant portion of a game win that can be attributed to a pitching performance. But it really is deeply flawed when so much hangs on how much the other 9 guys score. You're not going to win 20 games without being decent, but you can win 20 games when there are 20 better pitchers than you in the league.
Please make another stats video for specific sports like cricket and soccer and rugby
I would love to see a pathologist! love these videos dearly
Great video. Glad he actually tried to answer each question to the best of his availability instead of just saying words that don’t answer the question but do take up time in the video.
"the best measure we have for soccer performance for individuals, is goals." is an outrageous simplification of one of the most if not the most data-driven sports in the world.
This dude's knowledge checks out until they talk about the thing you know a lot about.
agreed, and this analyst certainly only works with the big three US sports nearly all the time.
I think he had a point. No matter how many data models the game produced, football is very subjective. How many times have you heard legends and former said Messi is the better player, but Ronaldo is the better goalscorer? Now they are saying Haaland is the better goalscorer while Mbappe is the more complete.
Outrageous. So defenders and midfielders always have bad performances according to how he measures performance. There are new stats like chances created, and possession value which are determining how scouting is done. This man does not know anything about soccer, he should have just said that.
@@boomboom1601 I think his point is that goal is the only objective reliable metric. A defender used to be measured by tackles, but now it became known that the better defender, the less tackles he had to make. The clean sheet can be attribute to the defensive midfielder or the goalkeeper or his partner or the tactics. A midfielder in a poccession-based team would make more passes than a counter-attack team, how do you define who is the better midfielder? We know Messi and Ozil has many assist, but they also provide the pre-assist that not many noticed.
This crop cutting is just too much. I’m getting dizzy yall.
Sports is a great way to get into statistics and programming. If you're into football/soccer and want to learn some programmings and stats, there's a good book on the topic. There's some serious statistical and spatial models used in football with great success!
Great video, disappointed no hockey though!
i feel enlightened!! this was a very cool interview for us non sports nerd types. Very awesome. Thank you Coach Oliver
If you nerd and don't play any sports you are double nerd not different type nerd
I stand corrected. :) @@Tomluke5823
I did stats for basketball in Australia a decade ago. We were taught to only give an assist to someone if their pass lead to an instant basket. So if you pass to someone who pump fakes and takes a couple dribbles then shoots that should not be an assist. I'd say majority of assists in the NBA are borderline assists.
Watching this vid is the closest I’ve ever gotten to understanding why people like sports
Hope they get Scott Steiner for the next Sports Math Support
Sènor Joe, the numbers don't lie, and they spell doom for you at Sackerfice.
I hope we get one for soccer as well
you mean football.
I obviously mean football. It's for the rest of the comments to get@@cktorm
yeah that was a super weak answer, felt like he doesnt rlly know much about the sport, which is understandable as its not big in usa
There are two popular sports with the same name, one of which has an acceptable alternative name. Why in a public forum, like a comments section for a video where both sports were discussed, would you insist on using the confusing same name, when you could just use the alternative name and clear up all confusion?
@@rokeYouuer because it's only confusing for americans. You are right about having an alternate name though, handegg seems appropriate.
The problem with pitcher wins is that pitchers don't win games all by themselves. Teams win and lose games, not individual players. The run support that a pitcher gets has a huge effect on his win percentage. The defense behind the pitcher also has an effect. So does the bullpen. A good pitcher on a bad team will probably win fewer games than a bad pitcher on a good team.
One of my favorite examples is Storm Davis with the 1989 Oakland A's. He had a record of 19-7 that season. He did this while allowing more runs than an average pitcher, despite the Coliseum being a good pitcher's park. His ERA was 4.36 in a league with an overall ERA of 3.88. His ERA+ was 85 (i.e. 15% below average). HIs WHIP was 1.506, while the league WHIP was 1.355. His FIP was 4.40, while the league FIP was 3.52. He got those 19 wins even though he pitched only 169.1 innings because the A's scored a lot of runs, had a good defense, and a great bullpen.
Bad fielders make slightly above average plays look spectacular cause they’re lunging and diving for balls that elite fielders make plays on routinely. That’s why Outs Above Average is an important metric.
It would be awesome to have such a video with more world famous sports like Football(soccer), cricket, boxing etc..!
The amount of data that sports collect is crazy!
Cricket has more numbers and stats than any other sport. Bring on Jarrod kimber onto the show for cricket
True shooting percentage puts a high value on made 3s, but doesn't similarly reflect missed 3s.
I feel so bad i thought he wasn't looking straight, always to the side and was getting annoyed, then i noticed the obvious.... Great insightful video!
Family lives in fairbanks, ak. Spent about a decade there. Can confirm, quite cold
I was expecting the answer to be: the free throw
This was super interesting, could be a pretty cool job but I imagine hard to get
How many decades did it take for people to recognize the offensive line affects a running back's production?
Baseball needs a new stat: Effective Batting Average. A single hit where the batter ends up stranded, with no other runners advancing is not an 'effective' at-bat. This new statistic could also be 'weighted' i.e. should a hit or at-bat be considered 'effective' if the hit comes in the ninth inning when the batter's team is behind by seven runs? (Ichiro, when he played for the Mariners had 1200 hits but only two walk-off hits (one homer, one single) in all those hits; he also hit a lot when there were two outs or Seattle was so afr behind it was not important).
Super interesting!
This guy must hit some crazy multis
Get This guy on again
Would have loved some Hockey stats mixed in here.
So true. I’ve been a hockey fan my whole life, but I’ve never really looked past total points and plus minus, which have some pretty large flaws when trying to gauge value
Hockey has some next level statistics that people just take as gospel. Really would like to see someone weigh in on CORSI and expected GF, GA stats
0:16 most efficient shot in basketball
4:21 defense vs offense: offense wins regular season, defense wins play offs
@TRP you need to do a video on Drew Griffith. He just ran a 3:57.08 1600m at the PIAA Track and Field Championships in Shippensburg, PA. This time is a new NFHS record and smashed Gary Martin's state record.
how did it take people that long to realize that almost all running backs' success is dependent on the offensive line. This is why Emmitt Smith was good. He ran through 10 feet wide holes that the GOAT offensive line created for him.
Bc back in the day rb’s actually ran through people. Yards after contact was higher just 10 years ago compared to today. And I’m sure the 15-20 years+ it was a higher number. You don’t see rb’s consistently trucking much any more. Back then it was a more violent league when rb’s were the hardest/most frequent hit position in the game. They deserved to get paid. Nowadays unfortunately, offenses aren’t dependent on the quality of an rb unless you’re Derrick Henry.
@@phomjachana If I had to guess I'd imagine RBs get much better yardage when they don't have to run people over. That, plus reduced risk of injury and reduced CTE damage, they're on the field more often, and for longer.
Was hoping they'd somehow slide in some sort of Emmitt Smith/Barry Sanders comparison question. If only breaking the defense's ankles was a statistic.
Well explained Dean.
was that the same "moving on" byte every time? xD
I want to talk to this guy all day
This has nothing to do with his lazy eye, but him always looking at whoever is next to the cameraman was so distracting
You could just feel him sweat with the soccer question
wide shot, medium shot, close-up, wide shot, medium shot, close-up, next question, repeat
I don't watch any of these sports...but that was great! I'm not even into math lol
Luv to see more of this guy regarding current team line ups, how/what to change...also any gambling tips ;)
Be like Moneyball 2
This guy has never seen a game of soccer in his life and it shows. But props on the rest
7:00 the pitcher win rule is so odd. it seems designed specifically for closers.
Interesting his take on Moneyball
Soccer is not at all a game of stats. The players that have the most impact are the ones that, in the first fase of construction, can find a way when they are on tight spot with heavy pressure. That generates more attacks and overall more offensive flow. And there's simply no stat for that.
What do you study? Andy Zaltzman studied Classics at Oxford. I think studying statistics is the thing to study
It’s tough to grade NFL and NBA players because there are so many variables. NBA players efficiency rating could take a hit based on the talent and spacing they have around them. Same with NFL players. Even individual advanced analytics that try to factor that in like expected rushing yards or yards after contact don’t account for oline play that can affect a rushers ability to gain downhill momentum b4 they are hit w contact. If Rhamodre Stevenson for example is hit behind the LOS bc of bad oline play, he’s going to be an inefficient runner vs a Raheem Mostert who has time and space to gain momentum. Nor does a statistic value their pass pro.
That’s why baseball is the perfect game for data nerds bc it’s very individual and there are way less variables
I hope he comes back.
Interesting I thought the most efficient shot is the long 2
a sports scientist what an innovation
Interesting I would have thought the most effective way to win games is build the better team
When I thought of a cool math job, I didn't think about this one. Interesting though
I thought the best shot in the nba was the layup/dunk, then free throws, the 3 pointers
The most efficient shot is the one that goes in.
LOVED this!!! Only questions is about the first question. Is a layup a more efficient shot vs a dunk? Or..... is a dunk really a "shot". Is this a definition thing? "Shot" is when a ball is released and has some level of trajectory? Whereas a "dunk" is literally placing the ball in the hoop, so while NOT 100%... it's pretty close. What is the rationale? Thanks anyone!
He's using the term layup to refer to any shot close to the basket, so a dunk would be included in that definition.
@@DellpodderExactly. That there are so many comments about this suggests he's maybe not that good communicating with sports people though
Why didn't I think about doing this when I signed up for college. Ugh.
Nobody
Michael: Shut up Toby..
😂😂😂
Most efficient position recently anything involving a Jokic low post
very interesting
What's a more important defensive stat, steals or blocks?
Nice vid
How did they not answer the questions about when is it better to punt in American football as opposed to going for a 4th down?
This guy looks exactly like a stereotypical statistician. Great vid!
He is wizards analyst lads, he knows squat
The gambling one is wrong. market making books no longer go for a 50/50 split. they usually have profiled their consumers very well that help them know which action is sharp vs recreational, which allows for a much more efficient closing line.
I hope the Pardon My Take Boys do something with this.
“He gets on base”
TS% could be defined simpler: average number of points added when the player takes a shot (whether he makes it or not, fouled on the shooting motion or not) divided by 2.
Note that he forgot to say it must be divided by 2.
He said the same thing, except he told us how to calculate the average number of points added per shot attempt
Can we PLEASE have him back?! I loved this!
you got it dude
The dunk is the most efficient shot
The goal stat in soccer was not one that was very useful until Messi and Ronaldo and is actually worse than evaluating basketball or football players based on the points they score, far worse. A better way to do would be impact on expected goals. I should write a paper
In Ben Simmons case the free throw pressure affected him 100% cause he stopped taking any by sitting out a whole season.
Ben shows up on a lot of highlight ... and lowlight clips. He was amazing when he was on, but man when he wasn't, it was awful.
4:00 Nick Anderson
Assists are complicated for another crucial reason! The player you pass it to has to make the shot!!!! A good passer on a bad team will never see those flashy passes on the stat sheet.
With that being said, the greatest passers should theoretically create good looks for teammates. Does it even out?
It should even out, but if a player is on a team with really bad shooters then they will get less assists, but it's not that many. The bigger factor is probably actually your teammates ability to get open and cut to the basket, because some players can do that a lot more than others, while a good 3 point shooter shoots like 5% higher than an average one
I understood some of what he said. The words that weren't sports or math.
QBs are the most depended position in football? If will they get devalued as well?
The dunk?