I have wandered down the road of Expected Points (ExPts) for hurling. This is a much vaunted way of judging the accuracy/finishing of players used mainly in soccer. Following from @dontfoul’s gaelic football model and using Dartfish’s excellent x,y capability I started down the road of tagging as many shots by location and by outcome. My model is very much in it’s infancy. I currently have only over 2,000 shots logged from club and Inter-County and will not be happy with a model until I reach 10,000 (sometime around Wexford’s next All Ireland!) – although the early signs are promising.
Looking at the 15 SHC games between the top 9 teams in hurling in 2017 gave me interesting results. I do caution about the system I use being in it’s early days but hopefully this will iron out in time.
Scoring Table – 2017 Season
This table takes the average Expected score for and against for each team in 2017. Galway are leaders in both categories. They generated the highest Expected Score For in their games (28.35) and had the lowest Expected Pts Against (20.85). A remarkable achievement to dominate both sides of the field.
Dublin’s season was a disaster from the sending off against Galway to the goal-fest Tipp put on in Semple Stadium. 2018 can’t come sooner. The numbers show their defence is where Pat Gilroy has to fix – conceding a massive 34.32 ExPts per game – 8 pts worse than the nearest team.
Tipperary’s failure to claim back to back titles appears to have come from a defensive lapse. Cork scored freely in Munster and Clare missed their chances to score on Tipp. The second highest ExPts Against of all nine teams is something a team aiming for an All Ireland cannot afford.
Scoring by Team
Above is a table showing how each team scored in relation to how the average shooter would perform in the same locations. This table shows us:
- Tipperary scored nearly 13 more pts than expected from the shooting areas they were in. Although a massive 10 pts of this was against Dublin where Tipperary ran riot.
- Waterford were nearly 8 pts above expected despite starting out against Cork with -3.96 pts on day one. They shot above expected for every game remaining.
- Most surprising is Kilkenny sitting bottom of the table. Their numbers for the year show that their forwards were completely off form this year. Ger Aylward, Chris Bolger and the great Richie Hogan all sit in the bottom 10 of 2017 shooting accuracy from play.
- The all conquering Galway team sit mid-table, yet beat all around. This comes down to their shots per game – Galway took on average 43.25 shots per game – 5 more than the nearest team! Galway played the long game and shot from distance, rarely driving inside the 21 (scoring 2 goals in the 4 games recorded and with no goal shots in either Final).
Conceding by Team
The table above shows how much a team conceded compared to what score we would expect them to have conceded from shots taken on them.
This might suggest Tipp having a good defence, as they conceded 5pts less than was expected, but the reality for the Premier County is that they were giving up a huge 39.5 shots per game (8 more per game than Galway) and survived by forcing some misses.
Kilkenny’s defence was only average, which by Cody’s standards is not good enough, but they can argue that they were not the weaker part of the team.
Above is a table of the players who scored more points from play than would be expected from the location of their shots.
John McGrath was the clear leader scoring over 8 pts above ExPts. He beat the ExPts in all 4 games recorded.
Kevin Moran and Jamie Barron were rightly recognised with All Stars and Player of the Year nominations and their shooting was of the highest standard. Both midfielders were in the top 5 most accurate shooters. Kevin Moran shot above Expected in every game after Munster.
Diarmuid O Keeffe is the only defender to make the top 10 and what a season he had, scoring in every NHL and SHC game he played for Wexford. Davy has made a menace out of the St Anne’s defender.
I hope to use the database of shots to work ahead on next season’s SHC, offering post game ExPts results. The data will build and should take weather and game state into account as it grows.
Any questions – then please comment.