With Irish raiders winning seven of the nine handicaps at last years Cheltenham Festival, the BHA announced a jumps rating review, and after some statistical analysis they found quite a significant inflation across almost all handicaps up to about 145 for hurdlers and 155 for chasers. From the graphs the BHA released it looks like ratings had inflated by around 6lb from averages of the period 2008 to 2011 compared to 2018 to 2021. Now I really have no idea how a competent handicapper could let this happen, it would be pretty basic stuff to keep levels around the same level year on year, but the problem is not isolated to the UK, as the Irish flat ratings have been inflating in a similar way for years, and they have countered that with sudden large drops at the end of the season. The obvious solution is to do the job properly in the first place and make sure that horses are dropping in the weights at a similar level to ones going up.
The UK handicap review did reach the conclusion that they needed to change their methodology, and decided to drop older/regressive horses quicker, and start certain sections of novice hurdlers off on lower marks. Their aim was to bring ratings back in line with the level they were at in the 2008 to 2011 period, and in a recent review they managed that with the hurdlers, but not with the chasers, with the hurdling graph now matching up very well with 08-11 levels which basically means a downwards shift of around 6lb in the ratings up to ratings around 145. The chase marks haven’t really gone down yet though although they still aim to bring them down in time.
Some interesting nuggets from the update shows the average rating increase for winning a handicap chase has gone from 6.16lb in the period 2017 to 2020 to just 5.45lb in 2020 to 2021 while over hurdles it’s gone from 5.95lb to 5.12lb in the same period. Horses finishing second have also seen a decreased rise by 0.81lb for hurdlers, and 0.514lb for chasers, while horses finishing third or worse have all seen increases in the amount they were dropped. All of this combined with a manual review of all ratings has contributed to an overall drop in the ratings, especially for hurdlers.
The review update then goes on to compare how Irish horses have done in UK handicaps in the prior 2017 to 2020 versus the 2021-2022 season, and it makes for interesting reading. From 2017-20 Irish hurdlers had an impact value of 1.28 in class 4 or higher handicap hurdles, while that has from a limited sample of just 112 runners this season, dropped to just 0.89. The first thing to note is the 2017-20 impact value of 1.28, which basically means they won 1.28 times the amount you’d expect given the field size, so if say in a sample of 10 runner races, you’d expect to win 10% of the time if every horse truly had an equal chance, so if you won 12.8% of the time you’d get an impact value of 12.8/10=1.28. Now I know many will look at that and say the Irish clearly had an advantage that needed sorting, but that’s totally ignoring so many factors. Look at any handicap and you’ll see that the betting can range from having an odds on favourite to horses that the market reckons have no chance, horses that are totally out of form and need to come down in the weights to have a chance. How many of this type do you think Irish trainers are going to go to the expense of sending over the Irish sea? Almost none I’d imagine. The distribution of horses being sent over to race in the UK will be totally different than the UK runners in the same races, and to compare them as like for like, as many have done, shows a poor grasp of even the basics of form analysis. The chase impact values were 1.0 in 2017 to 2020 and have dropped to just 0.84 this season from a sample of 74.
So to me the 1.28 IV looked like the handicapping between jurisdictions was fair for hurdlers and didn’t need changing, while the Irish were harshly treated over fences in the UK in the period 2017 and 2020. What else could you conclude when horses who will have consisted of a much higher percentage of in form horses (who win more often than out of form ones), who will have been likely working well, or else they wouldn’t send them, can only do as well as the UK horses, whose trainers won’t have been anywhere near as selective with running them. The sample size for the impact value drops to 0.89 for hurdlers, and 0.84 for chasers this season, are small, but the fact they have dropped is no surprise, as the UK horses have on the whole dropped in the weights, were as the Irish horses haven’t.
This leads us on the Cheltenham and of particular concern for Irish handicaps hurdlers at this years festival is that Kevin Blake found in his ATR Blog that the Irish entries for the non juvenile hurdlers have been given an average of a 4.9lb higher UK mark than their current Irish one. This compares to a 3.9lb higher mark last year, so despite the average level of ratings for the UK hurdling population having dropped this season, the handicapper still sees fit to raise the Irish horses more than last year. With logic like that it’s easy see how they ran into issues in the first place. The juvenile hurdlers have faired slightly better, they have received 3.3lb above their Irish rating this year versus 2.9lb last year. One thing to note is the BHA in it’s review pointed out that the median starting rating for novice hurdlers this season is 103 versus 108 in the period from 2017 to 2020, and this will have a knock on effect all the way up the scale, and in particular to the juveniles as they are all novices, so again the Irish ratings went up while the UK ones went down.
The average Irish chase marks handed out for Cheltenham have gone up 1.7lb from last year against their Irish marks, and while the handicapping review update found the ratings in the UK haven’t dropped in the same way the hurdlers have, there will still be some effect, there has to be when the average winners are getting raised less, and the losers dropped more, so the balance will have tilted toward the UK horses in the handicap chases too.
While the BHA review concentrated on how the Irish horses did against the UK in class 4 or better, I ran some figures myself in Proform for the past four years, for horses running in the UK that had their last run in Ireland, and included just class 1 and 2 handicap hurdles, to make it more relevant to Cheltenham, and as you’ll see from the screenshot above Irish horses really don’t have any advantage. The win IV is 1.09, but using the win and place IV is better as it increases the sample size, and that’s just 1.01, meaning they only had a fraction more places than would randomly be expected, and as previously stated given the profiles of horses crossing the water you would expect that to be higher if they were fairly handicapped.
I then ran the same criteria, except this time just for the Cheltenham festival, and it’s a very different picture, with a win IV of 1.91, and a win and place IV of 1.5, meaning we had 50% more places than you would randomly expect. You would also have made a healthy profit blindly backing such runner with 14 winners, when only 9.86 were expected given the Betfair SP, and 44 places when only 35.74 were expected.
The above is for class 1 and 2 handicap chases in the UK, again over the last four years for horses who had their last start in Ireland. It should be pointed out this can miss some Irish trained horses, for instance the Irish trained The Shunter won a handicap hurdle at Kelso last season prior to winning a handicap chase at Cheltenham, but for the purposes of reviewing handicaps this sample is just as good for showing which country their handicap mark was acquired in. It’s clear from the above that Irish chasers are poorly handicapped in the UK in recent times, and given this sample is over four years, so the recent review, which is ongoing, only affects a small portion of that data, it will only get worse for the Irish horses. The percentage of runners beaten is only 39.85%, while the win and place IV is only 0.9.
I then ran the same criteria but only including Cheltenham festival handicap chases, and while the picture is slightly better with a win IV of 1.54, you wouldn’t want to read much into that as the sample is small, and a better idea of how horses coming from a run in Ireland to a Cheltenham handicap do is looking at the percentage of rivals beaten which is just 44.96%, and the win and place IV which is 1.09, so certainly nothing in that to think the Irish chasers are advantageously treated.
It is stats like the hurdle ones that sparked the UK handicapping review, and while that review found there were gross errors in their methodology, the resultant fix will have penalized all Irish horses competing in the UK, when the data clearly shows it’s only Cheltenham that we overperform. So while the Irish horses were probably treated fairly over hurdles before the review the same couldn’t really be said for chases. With the Cheltenham stats being so much better than the overall, it also means if you took them out the horses who had their last run in Ireland did very poorly for the rest of the year in the UK.
So if we concentrate on horses who had their last run in Ireland and competed in a Cheltenham handicap hurdle. How come they do so much better than the overall figures for horses going to run in a UK handicap hurdle? I can think of a few reasons, but by far the biggest is that while having a winner at Cheltenham is a big deal for UK connections, it has always been a huge deal for Irish ones, maybe because for so many years they were very scarce. We love nothing better than going over there with a live chance, and just the anticipation of the party that would follow a winner, is enough for many to forego a chance of winning a big pot in Ireland in the preceding months, with everything instead geared towards Cheltenham. Ireland as seen from the graded results also has the best horses, and some of our trainers have great strength in depth, both mean they can be much more selective with the horses they target at the Cheltenham handicaps, and as such you would expect them to do much better than UK horses who for plenty, are just happy to be there.
Of course their is nothing wrong with targeting a handicap at the festival, UK trainers will target races too, some of them have adjusted to the Irish dominance at Cheltenham by targeting Aintree instead. So for the UK handicapper to adjust Irish horses ratings at Cheltenham to try even things up, is plainly wrong, and shows a total ignorance of the many reasons some horses win more handicaps than others. Better trainers win more handicaps, yet the UK handicapper has never penalized them domestically, could you say the same about the Cheltenham weights? A quick look at all the UK vs Irish ratings shows Gordon Elliot’s horses in particular have larger increases than the average. Better horses win more handicaps, mainly because flawed handicapping mean weight doesn’t count for as much as there methods allow for. In form horses win more handicaps, yet they’ve never made any attempt to even that out either.
I looked up the BHA website, and their description of their handicapping department is “The BHA has a team of eleven Handicappers whose job is to study and interpret the form. They publish a list of ratings every week based on performances on the racecourse”. That description doesn’t give them much scope to handicap trainers, or anticipate likely improvement from unexposed horses, or predict that a certain horse is being trained for a specific race, yet that is exactly what they seem to be doing in trying to level up results at a four day meeting. If the BHA or HRI have evidence of a horse not running on it’s merits, then they have processes to deal with it, it certainly isn’t the handicappers job, and maybe if they concentrated on doing their job they wouldn’t have to deal with grossly inflated ratings.
So to sum up I think the Irish handicap chasers next week have it all on, and despite the fact that they’ve done so well recently, and will likely have plenty laid out for the various handicap hurdles, the changes resulting form the UK handicap review could have a big impact, and the market on the whole probably won’t account for it, given last years results.