@wbagnr.bsky.social
Further analysis I'd like to do is see if the gini coefficient of points after 46 games correlates with position change. Is a more competitive league (by total points) likely to produce more movement than one with runaway leaders/relegation fodder?
November 10, 2025 at 1:52 PM
Further analysis I'd like to do is see if the gini coefficient of points after 46 games correlates with position change. Is a more competitive league (by total points) likely to produce more movement than one with runaway leaders/relegation fodder?
As Mark explains, the difference in one place can be large, and that's true in the Championship - 2nd vs 3rd and losing in the play-offs (Sheff Utd 👀) has a huge impact.
But I think this analysis largely tallies with the Premier League, except that mangerial changes can be a double-edged sword.
But I think this analysis largely tallies with the Premier League, except that mangerial changes can be a double-edged sword.
November 10, 2025 at 1:52 PM
As Mark explains, the difference in one place can be large, and that's true in the Championship - 2nd vs 3rd and losing in the play-offs (Sheff Utd 👀) has a huge impact.
But I think this analysis largely tallies with the Premier League, except that mangerial changes can be a double-edged sword.
But I think this analysis largely tallies with the Premier League, except that mangerial changes can be a double-edged sword.
I also checked the trend teams finishing +/- 1 place away, and repeated for 2 places and more than 6 places.
Generally there's been a reduction in teams finishing within 2 places (red line), until the 23/24 season when it rebounds back to pre-COVID levels.
Generally there's been a reduction in teams finishing within 2 places (red line), until the 23/24 season when it rebounds back to pre-COVID levels.
November 10, 2025 at 1:52 PM
I also checked the trend teams finishing +/- 1 place away, and repeated for 2 places and more than 6 places.
Generally there's been a reduction in teams finishing within 2 places (red line), until the 23/24 season when it rebounds back to pre-COVID levels.
Generally there's been a reduction in teams finishing within 2 places (red line), until the 23/24 season when it rebounds back to pre-COVID levels.
Below are the managerial changes and when the new boss took charge of their first match
Middlesbrough: Wilder -> Carrick (Gameweek 18)
West Brom: Bruce -> Coberan (GW18)
Norwich: D Smith -> Wagner (GW27)
Rotherham: Warne -> M Taylor (GW13)
QPR: Beale -> Critchley (GW23) -> Ainsworth (GW34)
Middlesbrough: Wilder -> Carrick (Gameweek 18)
West Brom: Bruce -> Coberan (GW18)
Norwich: D Smith -> Wagner (GW27)
Rotherham: Warne -> M Taylor (GW13)
QPR: Beale -> Critchley (GW23) -> Ainsworth (GW34)
November 10, 2025 at 1:52 PM
Below are the managerial changes and when the new boss took charge of their first match
Middlesbrough: Wilder -> Carrick (Gameweek 18)
West Brom: Bruce -> Coberan (GW18)
Norwich: D Smith -> Wagner (GW27)
Rotherham: Warne -> M Taylor (GW13)
QPR: Beale -> Critchley (GW23) -> Ainsworth (GW34)
Middlesbrough: Wilder -> Carrick (Gameweek 18)
West Brom: Bruce -> Coberan (GW18)
Norwich: D Smith -> Wagner (GW27)
Rotherham: Warne -> M Taylor (GW13)
QPR: Beale -> Critchley (GW23) -> Ainsworth (GW34)
The teams with the largest change in 22/23 after 15 games were:
Middlesbrough: +16
West Brom: +14
Norwich: -10
Rotherham: -11
QPR: -16
Each of these teams had a manager change which impacted their form severely (see next post)
Middlesbrough: +16
West Brom: +14
Norwich: -10
Rotherham: -11
QPR: -16
Each of these teams had a manager change which impacted their form severely (see next post)
November 10, 2025 at 1:52 PM
The teams with the largest change in 22/23 after 15 games were:
Middlesbrough: +16
West Brom: +14
Norwich: -10
Rotherham: -11
QPR: -16
Each of these teams had a manager change which impacted their form severely (see next post)
Middlesbrough: +16
West Brom: +14
Norwich: -10
Rotherham: -11
QPR: -16
Each of these teams had a manager change which impacted their form severely (see next post)
I conducted Spearman's Rank correlation analysis across the last 10 seasons, and barring two outliers, the correlation holds steady in the mid 0.7s, indicating a strong positive correlation.
The 19/20 season was obviously affected by COVID, but the 22/23 season is particularly odd.
The 19/20 season was obviously affected by COVID, but the 22/23 season is particularly odd.
November 10, 2025 at 1:52 PM
I conducted Spearman's Rank correlation analysis across the last 10 seasons, and barring two outliers, the correlation holds steady in the mid 0.7s, indicating a strong positive correlation.
The 19/20 season was obviously affected by COVID, but the 22/23 season is particularly odd.
The 19/20 season was obviously affected by COVID, but the 22/23 season is particularly odd.
They no longer have @iainmacintosh.bsky.social, they had their chance 😜
September 13, 2025 at 11:05 AM
They no longer have @iainmacintosh.bsky.social, they had their chance 😜
@norcrosscricket.bsky.social heard you talking on TMSA with Finny about leg byes going to the batter. I agree and hopefully this analysis helps in subsequent discussions!
September 4, 2025 at 9:50 AM
@norcrosscricket.bsky.social heard you talking on TMSA with Finny about leg byes going to the batter. I agree and hopefully this analysis helps in subsequent discussions!
So how does it compare to StrikeRate?
Similar but with some variance
Will Smeed - 157 SR, 2.175 RAR10
Phil Salt - 155 SR, 1.258 RAR10
So despite a similar SR, Smeed's runs have come off game-states where he's outperformed his peers more. This removes things like the PP from batting figures.
Similar but with some variance
Will Smeed - 157 SR, 2.175 RAR10
Phil Salt - 155 SR, 1.258 RAR10
So despite a similar SR, Smeed's runs have come off game-states where he's outperformed his peers more. This removes things like the PP from batting figures.
August 29, 2025 at 4:17 PM
So how does it compare to StrikeRate?
Similar but with some variance
Will Smeed - 157 SR, 2.175 RAR10
Phil Salt - 155 SR, 1.258 RAR10
So despite a similar SR, Smeed's runs have come off game-states where he's outperformed his peers more. This removes things like the PP from batting figures.
Similar but with some variance
Will Smeed - 157 SR, 2.175 RAR10
Phil Salt - 155 SR, 1.258 RAR10
So despite a similar SR, Smeed's runs have come off game-states where he's outperformed his peers more. This removes things like the PP from batting figures.
Runs above Replacement
That's the name I've used for the above metric described before. How many runs per ball better or worse than average is a batter vs all batters, accounting for which balls they faced, and how many wickets had fallen.
In the table, RAR10 means RAR per 10 balls; SR per 100.
That's the name I've used for the above metric described before. How many runs per ball better or worse than average is a batter vs all batters, accounting for which balls they faced, and how many wickets had fallen.
In the table, RAR10 means RAR per 10 balls; SR per 100.
August 29, 2025 at 4:17 PM
Runs above Replacement
That's the name I've used for the above metric described before. How many runs per ball better or worse than average is a batter vs all batters, accounting for which balls they faced, and how many wickets had fallen.
In the table, RAR10 means RAR per 10 balls; SR per 100.
That's the name I've used for the above metric described before. How many runs per ball better or worse than average is a batter vs all batters, accounting for which balls they faced, and how many wickets had fallen.
In the table, RAR10 means RAR per 10 balls; SR per 100.
This average can then be used as a baseline for every batter - given the game-state, did they do better or worse?
E.g.
The first ball of an innings will always be 0.1 - 0, and that goes for about 0.8 runs per ball. Ifa batter scores a single, then are +0.2. This can be summed for all balls
E.g.
The first ball of an innings will always be 0.1 - 0, and that goes for about 0.8 runs per ball. Ifa batter scores a single, then are +0.2. This can be summed for all balls
August 29, 2025 at 4:17 PM
This average can then be used as a baseline for every batter - given the game-state, did they do better or worse?
E.g.
The first ball of an innings will always be 0.1 - 0, and that goes for about 0.8 runs per ball. Ifa batter scores a single, then are +0.2. This can be summed for all balls
E.g.
The first ball of an innings will always be 0.1 - 0, and that goes for about 0.8 runs per ball. Ifa batter scores a single, then are +0.2. This can be summed for all balls
We can see from the graph that the Powerplay has a profound impact on run rate after the first set or two. Then there's a steady incline with the final 20 balls or so having a greater variability in run rate, which will come down to things like runs required and the higher number of wickets falling
August 29, 2025 at 4:17 PM
We can see from the graph that the Powerplay has a profound impact on run rate after the first set or two. Then there's a steady incline with the final 20 balls or so having a greater variability in run rate, which will come down to things like runs required and the higher number of wickets falling
So first I had to establish the average runs per ball, but I acknowledge that the number of wickets lost also contributes to the change in game-state. Being two down after 10 balls will have an impact on scoring vs not losing any wickets.
August 29, 2025 at 4:17 PM
So first I had to establish the average runs per ball, but I acknowledge that the number of wickets lost also contributes to the change in game-state. Being two down after 10 balls will have an impact on scoring vs not losing any wickets.
So to answer the questions, Extras are clearly valuable and some players seem to be able to generate extras while others don't. When matches are close, this could prove decisive.
//
I'm doing this as a way to improve my Python so if there's any questions or tips to I'm happy to discuss on here.
//
I'm doing this as a way to improve my Python so if there's any questions or tips to I'm happy to discuss on here.
August 26, 2025 at 12:37 PM
So to answer the questions, Extras are clearly valuable and some players seem to be able to generate extras while others don't. When matches are close, this could prove decisive.
//
I'm doing this as a way to improve my Python so if there's any questions or tips to I'm happy to discuss on here.
//
I'm doing this as a way to improve my Python so if there's any questions or tips to I'm happy to discuss on here.
Top 5 Players for Leg Bye Rate (per 100 balls) vs LBW Perc (vs Total Dismissals):
Z Crawley - 7.37 vs 9.09%
T Hartley - 6.86 vs 0.00%
R Whiteley - 6.48 vs 7.69%
S Narine - 6.03 vs 11.11%
L Wells - 5.84 vs 9.09%
Tom Banton - 1.67 vs 24.14% Ouch!
Z Crawley - 7.37 vs 9.09%
T Hartley - 6.86 vs 0.00%
R Whiteley - 6.48 vs 7.69%
S Narine - 6.03 vs 11.11%
L Wells - 5.84 vs 9.09%
Tom Banton - 1.67 vs 24.14% Ouch!
August 26, 2025 at 12:37 PM
Top 5 Players for Leg Bye Rate (per 100 balls) vs LBW Perc (vs Total Dismissals):
Z Crawley - 7.37 vs 9.09%
T Hartley - 6.86 vs 0.00%
R Whiteley - 6.48 vs 7.69%
S Narine - 6.03 vs 11.11%
L Wells - 5.84 vs 9.09%
Tom Banton - 1.67 vs 24.14% Ouch!
Z Crawley - 7.37 vs 9.09%
T Hartley - 6.86 vs 0.00%
R Whiteley - 6.48 vs 7.69%
S Narine - 6.03 vs 11.11%
L Wells - 5.84 vs 9.09%
Tom Banton - 1.67 vs 24.14% Ouch!
I feel Leg Byes are the extra where the batter has the most say, but do you get out LBW more?
Some players score leg byes without being given LBW, but the mix of batters and bowlers below suggests there's more analysis needed to understand which batters are most valuable for runs creation.
Some players score leg byes without being given LBW, but the mix of batters and bowlers below suggests there's more analysis needed to understand which batters are most valuable for runs creation.
August 26, 2025 at 12:37 PM
I feel Leg Byes are the extra where the batter has the most say, but do you get out LBW more?
Some players score leg byes without being given LBW, but the mix of batters and bowlers below suggests there's more analysis needed to understand which batters are most valuable for runs creation.
Some players score leg byes without being given LBW, but the mix of batters and bowlers below suggests there's more analysis needed to understand which batters are most valuable for runs creation.
At a player level, I wanted to see if extras varied by batter, or if the wider trends were consistent.
SR and ER have no correlation, so it's not just down to the bowler whether extras occur. Some batters strike well yet get few extras (H Klassen), while others can do both (Sams, Jordan).
SR and ER have no correlation, so it's not just down to the bowler whether extras occur. Some batters strike well yet get few extras (H Klassen), while others can do both (Sams, Jordan).
August 26, 2025 at 12:37 PM
At a player level, I wanted to see if extras varied by batter, or if the wider trends were consistent.
SR and ER have no correlation, so it's not just down to the bowler whether extras occur. Some batters strike well yet get few extras (H Klassen), while others can do both (Sams, Jordan).
SR and ER have no correlation, so it's not just down to the bowler whether extras occur. Some batters strike well yet get few extras (H Klassen), while others can do both (Sams, Jordan).
It's clear with only 2 fielders out, the risk of a No Ball (and subsequent Free Hit) dissuades bowlers during the Powerplay.
In the Death, Batters are likely to run more often on Leg Byes, but there will be fewer boundaries as more fielders can be outside the circle.
In the Death, Batters are likely to run more often on Leg Byes, but there will be fewer boundaries as more fielders can be outside the circle.
August 26, 2025 at 12:37 PM
It's clear with only 2 fielders out, the risk of a No Ball (and subsequent Free Hit) dissuades bowlers during the Powerplay.
In the Death, Batters are likely to run more often on Leg Byes, but there will be fewer boundaries as more fielders can be outside the circle.
In the Death, Batters are likely to run more often on Leg Byes, but there will be fewer boundaries as more fielders can be outside the circle.
Breaking Extras down, we can see trends across the three phases (sorry for the Excel graph)
Wides and Byes are generally consistent, Leg Byes and No Balls have opposite trends. Leg Byes represent 33% of extras, likely because it's rare to have a Fine Leg with only 2 men out, No Balls increase
Wides and Byes are generally consistent, Leg Byes and No Balls have opposite trends. Leg Byes represent 33% of extras, likely because it's rare to have a Fine Leg with only 2 men out, No Balls increase
August 26, 2025 at 12:37 PM
Breaking Extras down, we can see trends across the three phases (sorry for the Excel graph)
Wides and Byes are generally consistent, Leg Byes and No Balls have opposite trends. Leg Byes represent 33% of extras, likely because it's rare to have a Fine Leg with only 2 men out, No Balls increase
Wides and Byes are generally consistent, Leg Byes and No Balls have opposite trends. Leg Byes represent 33% of extras, likely because it's rare to have a Fine Leg with only 2 men out, No Balls increase
When looking at Runs off Bat vs Extras per 10 balls
Overall - 13.26 / 0.85
Powerplay - 12.69 / 0.95
Middle - 12.70 / 0.74
Death - 15.28 / 1.01
So the death overs see teams get more Extras, but as the Run Rate is so much higher it forms a lower % than the Powerplay.
Overall - 13.26 / 0.85
Powerplay - 12.69 / 0.95
Middle - 12.70 / 0.74
Death - 15.28 / 1.01
So the death overs see teams get more Extras, but as the Run Rate is so much higher it forms a lower % than the Powerplay.
August 26, 2025 at 12:37 PM
When looking at Runs off Bat vs Extras per 10 balls
Overall - 13.26 / 0.85
Powerplay - 12.69 / 0.95
Middle - 12.70 / 0.74
Death - 15.28 / 1.01
So the death overs see teams get more Extras, but as the Run Rate is so much higher it forms a lower % than the Powerplay.
Overall - 13.26 / 0.85
Powerplay - 12.69 / 0.95
Middle - 12.70 / 0.74
Death - 15.28 / 1.01
So the death overs see teams get more Extras, but as the Run Rate is so much higher it forms a lower % than the Powerplay.
Extras constitute 6% of all runs scored across matches in the data, but this varies by phase of game:
Powerplay - 6.7%
Middle - 5.5%
Death - 6.2%
It's interesting that Powerplay has the highest % of Extras - we'll look at why later on.
Powerplay - 6.7%
Middle - 5.5%
Death - 6.2%
It's interesting that Powerplay has the highest % of Extras - we'll look at why later on.
August 26, 2025 at 12:37 PM
Extras constitute 6% of all runs scored across matches in the data, but this varies by phase of game:
Powerplay - 6.7%
Middle - 5.5%
Death - 6.2%
It's interesting that Powerplay has the highest % of Extras - we'll look at why later on.
Powerplay - 6.7%
Middle - 5.5%
Death - 6.2%
It's interesting that Powerplay has the highest % of Extras - we'll look at why later on.