Great analysis, Benn! Since this is purely statistical, I think it misses three big themes: the pressure in big tournaments such as the Euros and the World Cup, the historical performance of national teams, and the culture of football.
1) The pressure during the WC is so high that even great players can miss penalty kicks, such as Jorginho in the final at Euros 2020 (held in 2021) or Kane against France in the quarter-final. I wonder if there's a distribution that can include "pressure" and adjust the penalty taker's conversion rate. In that case, I doubt that even PK specialists will have a 95% conversion rate especially in the knockout stage, and it likely will have a steep fall given the stage of the knockout game (R16, QF, SF, F).
2) There are some teams that have historically been bad at taking penalties, such as England and Spain, despite having an array of talented players and practiced penalties before big games (e.g. Spain before the QF with Morocco, England before the F with Italy in Euros 2020). Despite their efforts to improve this obvious deficiency with their prodigious talents, these teams have failed time and again. In other words, I would expect a Spanish or English PK specialist may have a different distribution than a French PK specialist, so PK specialists may make a bigger difference for some teams than others.
3) Lastly, given the culture of football, it may be more acceptable to have one goalkeeper out of the three be a specialist at saving penalties than having 1-2 PK specialists. That may arguably make a bigger difference, such as Bono did for Morocco against Spain, or Martinez for Argentina against France.
1. ) On the pressure, sure, but it seems like that'd apply to non-specialists just as much (which you see, because PK rates go from 80-85% outside of the WC to 70% in it). Maybe 95% is high, but it doesn't seem at all crazy to think that practicing these kicks way more wouldn't make people considerably better at them. (Honestly, though, I think the harder part of this is that it's almost impossible to say if it works. You can imagine a team doing this, and the entire effort is then assessed against a single WC penalty. If they miss, people would be like, it doesn't work, even if they would make the next 99. It's such a low frequency event it's tough to play the numbers).
2.) Fair, though that makes me wonder why? Is it just a psychological thing, that teams that are bad at it feel more pressure? Or is it just a statistical artifact that some teams are inevitably going to have a string of bad luck, and then we add narratives to that after the fact? It'd be really strange, for instance, if an NBA team couldn't shoot free throws, even if all their players were perfectly fine free throw shooters when they played for other teams.
3.) Yeah, someone else mentioned this, and it seems like a potentially better idea actually. You have to carry the goalkeepers, most of them never play anyway, they get five chances vs one, and it's probably slightly less of a crapshoot than taking the PK. https://twitter.com/kevleong/status/1606376561959002112
Interesting post. A couple of thoughts on the core of your post:
– Teams have in the past made those specific substitutions in the 118th/119th minute to bring on better penalty takers (works sometimes but many times to not great outcomes such as England in the Euros final), but mental fortitude plays a big role, so realistically, I don't think 95% rates may be achievable. The very best penalty takers (who albeit also have other skills) are at ~90% (Jorginho, Fernandes, etc), and I don't think specialists per se would surpass that.
– 2 seems a bit high for the number of substitutions teams may be able to hold back for the penalty specialists. For example, as you said, of the 17 substitutions, only 7 took a penalty. Others were brought on to influence the last 15 minutes in extra time. In most extra time periods, team were a goal down in the last 15 and needed to score (Croatia vs Brazil, France vs Argentina) or defending for their lives (Morocco vs Spain, Netherlands vs Argentina). Not being able to make substitutions to help that cause may have led to them losing the game prior to penalties.
– As an aside, in regular play, a Penalty in general is supposed to be an extreme punishment for a foul in the box to force defenders to be more careful. Sure, you could require the player who was fouled to take it which would say reduce conversions from ~80% to 60-70%, but the point really is to make it extreme so defenders don't commit fouls. The typical game sees about 0.25 penalties per game, although this has gone up with the introduction of VAR, but on net the impact of the rule change you mention would only be about 0.05xG, which is not huge relative to the typical 2.25-2.5xG of a game.
– Lastly, this analysis is for the World Club and perhaps there is a role for specialists to influence the outcome more in such tournaments, but these players likely would not find themselves in a job at club level. In a club team's season, maybe 2-3 of there ~70 games a club plays go to shootouts (because over half their games are in a league format), so it's hard to see a room in the squad for someone who can literally only take penalties in a shootout. Given that, it's unlikely that very many players would want to be of this ilk, since they're likely only employed at the national level.
- On one hand, agreed that 95% would be very high in today's game. But on the other hand, those numbers are from players who aren't entirely focused on it. You see this a bit with field goals in the NFL (https://www.footballperspective.com/field-goal-rates-throughout-nfl-history/) or with how good people have gotten at 3 pointers once they became such a focus. Granted, teams would lose that edge pretty quickly if it works, but 1) do it first, and 2) if that happens, maybe that means they'd get rid of PKs.
- 2 might be high, but 1 is definitely possible. 40% of teams had unused subs, and a handful of the subs were made on the 120th minute.
- Fair. I think my bigger gripe with anyone taking them is how it affects stat lines. At a minimum, I don't think those goals should count toward people's goal totals, just as shootout goals don't.
- Yeah, I agree this probably don't work in leagues. And for players at that level, they probably wouldn't be willing to commit to being a PK specialist. But what about players who have no other option? There are people who are willing to play for years in low-A baseball, or arena league football. I'm sure there are lots of people who are great-but-not-elite soccer players who'd happily commit to taking lots of PKs if it got them on a roster.
What if I told you that we have already seen a specialist in penalty shootouts at a World Cup? But it wasn't a penalty taker, but a goalkeeper! In 2014, the Netherlands subbed a goalkeeper (Tim Krul) at the very end of overtime. It worked out since he saved two penalties. Here is a vid about it: https://www.youtube.com/watch?v=SP2IkwrLbu4
Even though it was a success, it was very controversial at the time because soccer only allowed had 3 subs for the whole 120 min, and some argued that by saving a sub for the GK the coach actually hindered his team chances to win in regulation.
A few other people suggested this, and it makes a lot of sense, especially with the bigger rosters and extra subs. Teams have to keep 3 goalkeepers on their roster now, and only 6 (of 32) teams played more than 1. So why not have one that's especially good at PKs? And if you've got the sub towards the end of extra time, use it.
Benn hi! Could you elaborate further on ```If both teams have a lineup without a specialist, the final outcome is a coin flip, with each team winning half the time``` ?
You mean on average, assuming 50% change of every kicker + goalkeeper of making / saving the shot?
Sure - I meant if assume both teams have equally talented penalty takers, they'll each have a 50-50 shot of winning. Obviously, that's not actually exactly true, but for the sake of estimating how much it helps to add one (or more) really good kickers, it seemed fair to assume that the two teams start off evenly matched.
More specifically, I assumed each team had 5 kickers: The first two made their shots 75% of the time, the next one made 70%, and the last two made 65%. Then, when adding specialists, I'd add them to the top of the lineup, and bump the worst kicker off.
Soccer is the only sport I know where the referee jersey (kit) colors in a given game are determined by the jersey colors of the teams playing. There are typically five options: black, red, yellow, green and blue.
No other league or sport is like this as far as I can tell, where the teams on the field and their jersey choices affect how the referees will dress on a given day.
I think it says something about why there is so much flopping.
Sometimes you gotta call it out like Schwarber did. 9th inning, he's a power hitter, you gotta put the line in the sand like he did. He's a Top 10 batter in the league right now for not putting up with shenanigans, not to mention his hard HRs.
Hernandez is in front of Philly fans and gets booed as he wipes the plate. You can't get any better than this. It's awesome.
Great analysis, Benn! Since this is purely statistical, I think it misses three big themes: the pressure in big tournaments such as the Euros and the World Cup, the historical performance of national teams, and the culture of football.
1) The pressure during the WC is so high that even great players can miss penalty kicks, such as Jorginho in the final at Euros 2020 (held in 2021) or Kane against France in the quarter-final. I wonder if there's a distribution that can include "pressure" and adjust the penalty taker's conversion rate. In that case, I doubt that even PK specialists will have a 95% conversion rate especially in the knockout stage, and it likely will have a steep fall given the stage of the knockout game (R16, QF, SF, F).
2) There are some teams that have historically been bad at taking penalties, such as England and Spain, despite having an array of talented players and practiced penalties before big games (e.g. Spain before the QF with Morocco, England before the F with Italy in Euros 2020). Despite their efforts to improve this obvious deficiency with their prodigious talents, these teams have failed time and again. In other words, I would expect a Spanish or English PK specialist may have a different distribution than a French PK specialist, so PK specialists may make a bigger difference for some teams than others.
3) Lastly, given the culture of football, it may be more acceptable to have one goalkeeper out of the three be a specialist at saving penalties than having 1-2 PK specialists. That may arguably make a bigger difference, such as Bono did for Morocco against Spain, or Martinez for Argentina against France.
1. ) On the pressure, sure, but it seems like that'd apply to non-specialists just as much (which you see, because PK rates go from 80-85% outside of the WC to 70% in it). Maybe 95% is high, but it doesn't seem at all crazy to think that practicing these kicks way more wouldn't make people considerably better at them. (Honestly, though, I think the harder part of this is that it's almost impossible to say if it works. You can imagine a team doing this, and the entire effort is then assessed against a single WC penalty. If they miss, people would be like, it doesn't work, even if they would make the next 99. It's such a low frequency event it's tough to play the numbers).
2.) Fair, though that makes me wonder why? Is it just a psychological thing, that teams that are bad at it feel more pressure? Or is it just a statistical artifact that some teams are inevitably going to have a string of bad luck, and then we add narratives to that after the fact? It'd be really strange, for instance, if an NBA team couldn't shoot free throws, even if all their players were perfectly fine free throw shooters when they played for other teams.
3.) Yeah, someone else mentioned this, and it seems like a potentially better idea actually. You have to carry the goalkeepers, most of them never play anyway, they get five chances vs one, and it's probably slightly less of a crapshoot than taking the PK. https://twitter.com/kevleong/status/1606376561959002112
Interesting post. A couple of thoughts on the core of your post:
– Teams have in the past made those specific substitutions in the 118th/119th minute to bring on better penalty takers (works sometimes but many times to not great outcomes such as England in the Euros final), but mental fortitude plays a big role, so realistically, I don't think 95% rates may be achievable. The very best penalty takers (who albeit also have other skills) are at ~90% (Jorginho, Fernandes, etc), and I don't think specialists per se would surpass that.
– 2 seems a bit high for the number of substitutions teams may be able to hold back for the penalty specialists. For example, as you said, of the 17 substitutions, only 7 took a penalty. Others were brought on to influence the last 15 minutes in extra time. In most extra time periods, team were a goal down in the last 15 and needed to score (Croatia vs Brazil, France vs Argentina) or defending for their lives (Morocco vs Spain, Netherlands vs Argentina). Not being able to make substitutions to help that cause may have led to them losing the game prior to penalties.
– As an aside, in regular play, a Penalty in general is supposed to be an extreme punishment for a foul in the box to force defenders to be more careful. Sure, you could require the player who was fouled to take it which would say reduce conversions from ~80% to 60-70%, but the point really is to make it extreme so defenders don't commit fouls. The typical game sees about 0.25 penalties per game, although this has gone up with the introduction of VAR, but on net the impact of the rule change you mention would only be about 0.05xG, which is not huge relative to the typical 2.25-2.5xG of a game.
– Lastly, this analysis is for the World Club and perhaps there is a role for specialists to influence the outcome more in such tournaments, but these players likely would not find themselves in a job at club level. In a club team's season, maybe 2-3 of there ~70 games a club plays go to shootouts (because over half their games are in a league format), so it's hard to see a room in the squad for someone who can literally only take penalties in a shootout. Given that, it's unlikely that very many players would want to be of this ilk, since they're likely only employed at the national level.
- On one hand, agreed that 95% would be very high in today's game. But on the other hand, those numbers are from players who aren't entirely focused on it. You see this a bit with field goals in the NFL (https://www.footballperspective.com/field-goal-rates-throughout-nfl-history/) or with how good people have gotten at 3 pointers once they became such a focus. Granted, teams would lose that edge pretty quickly if it works, but 1) do it first, and 2) if that happens, maybe that means they'd get rid of PKs.
- 2 might be high, but 1 is definitely possible. 40% of teams had unused subs, and a handful of the subs were made on the 120th minute.
- Fair. I think my bigger gripe with anyone taking them is how it affects stat lines. At a minimum, I don't think those goals should count toward people's goal totals, just as shootout goals don't.
- Yeah, I agree this probably don't work in leagues. And for players at that level, they probably wouldn't be willing to commit to being a PK specialist. But what about players who have no other option? There are people who are willing to play for years in low-A baseball, or arena league football. I'm sure there are lots of people who are great-but-not-elite soccer players who'd happily commit to taking lots of PKs if it got them on a roster.
Great article Benn!
What if I told you that we have already seen a specialist in penalty shootouts at a World Cup? But it wasn't a penalty taker, but a goalkeeper! In 2014, the Netherlands subbed a goalkeeper (Tim Krul) at the very end of overtime. It worked out since he saved two penalties. Here is a vid about it: https://www.youtube.com/watch?v=SP2IkwrLbu4
Even though it was a success, it was very controversial at the time because soccer only allowed had 3 subs for the whole 120 min, and some argued that by saving a sub for the GK the coach actually hindered his team chances to win in regulation.
A few other people suggested this, and it makes a lot of sense, especially with the bigger rosters and extra subs. Teams have to keep 3 goalkeepers on their roster now, and only 6 (of 32) teams played more than 1. So why not have one that's especially good at PKs? And if you've got the sub towards the end of extra time, use it.
Such great analysis! You are the rain man of soccer/football statistics. 🔥🔥🔥🔥⚽⚽⚽⚽⚽
Benn hi! Could you elaborate further on ```If both teams have a lineup without a specialist, the final outcome is a coin flip, with each team winning half the time``` ?
You mean on average, assuming 50% change of every kicker + goalkeeper of making / saving the shot?
Sure - I meant if assume both teams have equally talented penalty takers, they'll each have a 50-50 shot of winning. Obviously, that's not actually exactly true, but for the sake of estimating how much it helps to add one (or more) really good kickers, it seemed fair to assume that the two teams start off evenly matched.
More specifically, I assumed each team had 5 kickers: The first two made their shots 75% of the time, the next one made 70%, and the last two made 65%. Then, when adding specialists, I'd add them to the top of the lineup, and bump the worst kicker off.
Soccer is the only sport I know where the referee jersey (kit) colors in a given game are determined by the jersey colors of the teams playing. There are typically five options: black, red, yellow, green and blue.
No other league or sport is like this as far as I can tell, where the teams on the field and their jersey choices affect how the referees will dress on a given day.
I think it says something about why there is so much flopping.
Baseball umpires, by contrast, wear this....
https://www.google.com/url?sa=i&url=https%3A%2F%2Fnews.sportslogos.net%2F2021%2F10%2F09%2Fexplaining-the-ftx-patch-worn-by-mlb-umpires%2Fbaseball%2F&psig=AOvVaw0IiQCr_HCZoLKap6BfWzrK&ust=1671912582846000&source=images&cd=vfe&ved=0CA8QjRxqFwoTCJjp2-7FkPwCFQAAAAAdAAAAABAF
Nothing like watching Angel Hernandez call a strike that was in the dirt while he's wearing the FTX.
all time angel hernandez moment: https://twitter.com/umpjob/status/1518412755153264640
Sometimes you gotta call it out like Schwarber did. 9th inning, he's a power hitter, you gotta put the line in the sand like he did. He's a Top 10 batter in the league right now for not putting up with shenanigans, not to mention his hard HRs.
Hernandez is in front of Philly fans and gets booed as he wipes the plate. You can't get any better than this. It's awesome.