Ok, so I was right; it's a massive risk then. Hmmm............not sure what way to go now. Was planning on £5 e/w on the following; Beshabar (33-1), Cappa Bleu (25) Ballabriggs (33) Imperial Commander (33) T43 (20) Poker Desivola (50) Oscar Time (50) and Prince Bauchene (12).......................anyone wanna put me off any of these as particular risk of being non-runners?
Ok, so I was right; it's a massive risk then. Hmmm............not sure what way to go now. Was planning on £5 e/w on the following; Beshabar (33-1), Cappa Bleu (25) Ballabriggs (33) Imperial Commander (33) T43 (20) Poker Desivola (50) Oscar Time (50) and Prince Bauchene (12).......................anyone wanna put me off any of these as particular risk of being non-runners?
My advise would be, don't bother going e/w in the ante post market. Too many uncertainties to be doubling up your stake at this early stage IMO. (I never bet e/w in the National - too many runners. I'd rather back another fancy than go e/w)
Peanuts - if I bet on a horse now (to get what seems a good price), do I lose my stake money if the horse is a non-runner? Something tells me I do, unlike footie where betting on a non-playing 'first goalscorer' results in a non-bet/stake refund. I'm just weighing up the best option as I am aware that many planned runners don't end up running! Please advise.
Agree with ISLS and AA. Cheers chaps. Prices don't drop for all runners, especially on the Exchanges. Dont Push It drifted way out to 90 or so on Betfair before the market got wind that McCoy might ride him (from memory about 2 weeks before the race) and Mon Mome's price was never longer than at the Off in 09 but if you fancy a horse that is a definite intended runner and is liable to attract interest (eg those that Walsh and McCoy may opt to ride - Dont Push It went off 10/1 Jt Fav!) then ante-post makes sense. For e/w betting I would wait until at least 5 places are offered. Betting strategy is a personal matter but, for my GN system, I always back the 4 or 5 selections e/w with the bookies (eventually) and use the Exchange Win market as a trading/positioning vehicle. If you've got say 4 running for you, it's not inconceivable that 3 may pay off (blimey it happpened for my system in 2008 and 2010) and I'd do some serious damage to the TV if I'd backed them all to win and they came in 2,3,4.
Blimey, there will be a lot of unhappy ante-post punters that got stuck in after the weights were announced. I had a little position myself and was sitting on a nice profit (that'll teach me to get greedy) ( Weights will go up at least 2lbs and Albertas Run presumably not a certain participant. Puts the cat among the pigeons. Wouldn't hurt T43 unduly as regards stat-profile, in fact could assist in pushing others further above 11.00.
Hab SoSlI' Quch! All the runners will carry at least 2lbs more weight (maybe more if Albertas Run, the new topweight, doesn't run) but T43 is still good.
Peanuts, as much as I love you after my little win on the Welsh National (thanks again) that is just Clingon to me mate!
Keep up, what Peanuts said was ...Weights will go up 2lbs,Albertas Run might not run,cat is amongst Pigeons . and T43 wont be hurt, and Ante aint gonna be happy. What is there not to understand ???
There is a good G N trial tomorrow @ Haydock. I'll be interested to see how Lively Baron and Viking Blond perform over 3 and 1/2 miles in heavy ground, as well as Tea For three and Monbeg Dude
Very impressive indeed by Cappa Bleu with a lot more weight than last year. Last Welsh-trained winner of the GN was 108 years ago but there's 2 bloody good chances this year. At least I hope there's 2 chances - an anxious wait for confirmation that T43's OK after his 43L 10th. Jumped and travelled well enough for the first 3m but maybe the notorious "Haydock Heavy" did for him in the last 1/2 mile. As you'd expect, Scholfield wasn't particularly hard on him once the chance to challenge went. On paper not unduly concerned about the result. Mon Mome only managed a 42L 7th in the same race in 2009 before romping home at Aintree (with an even worse run in the Midlands National in between). Further back Earth Summit won the GN in 1998 after a 46L 5th in the Haydock GN Trial............maybe a good omen ;o)
While the epithet "Grand National trial" (appended overnight by some goon writing for BBC Sport online, who seems to think that Papillon was trained by Willie Mullins) may be over-the-top for the Bobbyjo Chase (Hedgehunter did indeed win both BJ and GN in the same season but the GN performance of other BJ winners that have attemped it read: UR / PU / F / UR / 6th), nonetheless, as the main 3m chase at home for Irish contenders to be run after the GN weights have been set, it is an interesting guide as to current form for those with GN credentials. For example, both Black Apalachi and Oscar Time ran well in the BJ before coming home runners-up in their same season GNs. Notably, Black Apalachi also had a Becher Chase win to his name and had run very well in the prior GN and Oscar Time had been a close runner-up on heavy in the Irish National a year earlier. Only Snowy Morning (3rd in both races in 2008) ticked neither stamina nor GN fence boxes prior to his GN placing. So, what about this year's race, in which all 6 runners had GN entries? Roi Du Mee won the BJ yesterday but has pretty dismal GN stats - in 22 chases he's never won a proper contest at any trip on decent ground and his only attempt at beyond 3m1f was a very remote 8th in last year's Irish National on Good-to-Soft. He has to carry at least 11.03 if he lines up at Aintree on 6th April and my model gives him little chance of being in the shake-up. Price Du Beauchene ran another nice race and came in a close 2nd yesterday. While he has done little wrong since his move to Ireland and has obvious staying credentials in his family, he's never been tested beyind 3m1f and ticks very few GN boxes. He will have to carry at least 11.05 around Aintree (if Albertas Run comes out it will be at least 11.07). Don't Push It and Neptune Collonges of course carried 11.05 and 11.06 to win recent GNs but both had quality form at 3.5miles+. Like PdB, Snowy Morning had no form beyond 3m1f but at least his OR for the GN was 10lbs below his 155 mark when placed in the BJ. PdB has no such advantage. Seabass did his 2013 GN stats no harm with his close 3rd yesterday and is a strong contender to be in the shake-up once again. However, his GN stats remain inferior to both Teaforthree and Cappa Bleu on any going but particularly if it were to come up soft. According to my model, a strong run but a place is his best prospect, barring others' misfortune. Neither Rare Bob nor Bob Lingo should trouble the Aintree judges and, sadly, on stats aswell as form, Oscar Time is not the horse he was when a gallant 2nd to Ballabriggs 2 years ago. This side of the Irish Sea, the Racing Plus Chase at Kempton was most notable for the truly awful jumping of Wyck Hill, apparently caused by a cut to a hind leg. He would need to do something special at Cheltenham to put his stats anywhere near giving him a chance to contend at Aintree. Otherwise, despite creditable efforts, there was little of relevance to the GN prospects of Quinz or Roberto Goldback, which are not great in either case. It was a shame that the Eider Chase was snowed off. On the stats, a win or near-thing would have propelled 150/1 Tarquinius into the reckoning if the GN going were to come up soft~heavy. While he has entries at the Festival, it's possible Elliott may route him to Uttoxeter for the Midlands National in 3 weeks time; but he must win or go close on soft or softer there to give the required boost to his GN [soft~heavy] stats. In any event, there remains another funky outsider for Good-to-Soft or softer going at the GN - more anon. First forfeit stage on Tuesday.
Had Roi Du Mee at 6-1. I take it that your ruling out due to distance? Five wins and a second in last 6 races ain't bad.
Congrats on RdM on Saturday. I was going to be on Quito De La Roque – shame they just can’t keep him sound.
THE MODEL
A word of explanation about my model may be helpful. It has 2 important distinguishing features compared to most stat- or trend-based GN analysis. Firstly, with one exception (any 7 year-old), my stat-model is not a process of elimination. It doesn't say, for example, no top-weight has won for 39 years so we’ll eliminate all top-weights. Rather, considering a variety of criteria, it applies penalties and credits to each runner based upon the degree of "mis-fit" or "strong-fit" of their career stats to the corresponding pre-GN stats of successful runners in all GNs since 1988 (excl the Red Marauder GN). So, being top-weight incurs a penalty but it is possible (difficult but definitely possible) that credits achieved for other criteria more than compensate for that penalty and give a top-weight a winning profile. Secondly, the “successful runners” we’re looking to as statistical yardsticks are not just the GN winners but all horses placed within 4 lengths of the winners since, but for the vagaries of fortune, they could easily have won (4L is nothing over a 4.5 mile steeplechase with a big field). As an aside, in fact, the model takes account of all horses finishing within 30L of the winners, since their strengths and weaknesses and composite stat-scores (decreasing with distance from the winner) all help to inform the model’s algorithm. When the model is tweaked (after every GN) the aim is to reconcile unexpected results but also to ensure robustness. So, all tweaks to the algorithm are back-tested against all GNs since 1988 to confirm consistency of stat-scores vs distances for all horses winning and within 30L of the winners (of course allowances are made subjectively for meaningful mishaps during running). The 2 distinguishing features of the model matter because they are the reason that it has (more or less) been able to cope with evolving trends in the race. There are two particular examples where the model has identified a GN winner that was eliminated by many trend-followers: 1. Prior to 2009, many routinely eliminated all French-bred runners because no French-bred had ever won the GN. But that overlooked the fact that 4 French-breds had gone very close indeed to winning in recent years: Encore Un Peu in 1996, Mely Moss in 2000, Clan Royal in 2004 and Royal Auclair in 2005 all finished 2nd, the first two by as little as 1.25 lengths. Then Mon Mome came in at 100/1 – lovely jubbly. 2. Prior to 2010, many routinely eliminated all runners carrying more than 11.01, since no GN winner had carried more for over 25 years. However, that overlooked the fact that 5 horses carrying 11.06+ had been GN runner-up since 1998 (two were top-weights). Increasing compression of the handicap had also led to the first 4 home in 2009 carrying 11.00+. Then Don’t Push It came in carrying 11.05 (followed of course by the French-bred Neptunes Collonges carrying 11.06 in 2011 - alas on this occasion the model preferred Sunnyhillboy!).
ROI DU MEE
You’re spot on about consistency of form this season and this is one of the criteria used by the model, for which he scores the maximum credit. He gets an additional modest credit for winning the Bobbyjo. However, he has several negatives. He is modestly penalised for carrying 11.03 (that will get worse if the weights rise of course) and, again, for never having won a chase at a major left-handed track. He is more significantly penalised because in 22 chases he has never won a proper contest on ground officially better than soft and never won any race (chase or hurdle) in a time equivalent (according to my model’s calculations) to Good-to-Soft or better. In 5 chases on (officially) Good or Yielding he has never been placed. That’s a big problem if the GN is run on anything better than Good-To-Soft because he will (according to my model) struggle to go the pace. That penalty would be partially mitigated if he were to go close on decent going in the 2.5m Ryanair at the Festival, for which he has an entry, but it would need to be a 3m chase to remove it altogether. One might say, he’ll be suited by soft going at Aintree as he’s done all his winning on Soft or Hvy. But in order to score as GN Soft Ground contender, the model also requires as a yardstick of the required stamina that a runner has won or performed creditably over at least 3.5 miles. The winners of all 7 GNs run on Good-to-Soft, Soft or Heavy ground (officially described or time-adjusted and excluding the Red Marauder GN) since 1988 ticked this box and only 1 horse finishing within 10L of the winners failed to do so (Garrison Savannah which had won the Gold Cup 3 weeks before and carried 13lbs less in the GN). Unfortunately RdM has done all his winning at no further than 3m1f. His only attempt at further was on officially Good-to-Soft in the Irish National (the winning time suggests that it was softer which should have suited) in which he was always behind and finished a remote 8th. Again, that could change if Elliott gives him a run in the 4.25m Midlands National in 3 weeks time, particularly if it were to be on soft or softer. However, there’s another negative stat for him on Soft~Heavy because he is an 8 y-o. In the 5 GNs since 1988 (again excl the Red Marauder GN) which were run on officially soft or heavy, or whose time suggested that the ground was such, no horse younger than 9 has made the frame. Indeed, of the 27 horses younger than 9 to run in these 5 GNs, only 4 finished (14.8%) [none within 30L of winner] compared with an average finishing rate of 22.8% for these races. Having said all that, he’ll probably bolt up!
Apologies for the length of this – hope it hasn’t sent you to sleep.
They've been adjusting the fences pretty much annually for some years (gradually levelling the landing areas at the likes of Bechers for example). As I understand it, this year they've made the cores of the fences plastic rather than wooden (that may be a bit of a simplistic description) but haven't reduced the dimensions of the fences so I doubt that they'be been made easier to jump, just less injurious (hopefully) to those that fail to clear them. In any event, you can only model the stats of actual races. While my system does seek to accomodate evolving trends it can only do so based upon actual results, with often the biggest clues to new factors coming by means of the placed horses. The other change this year is to reduce the distance by a furlong (200 yards). May not sound much but I could have done with even 2 yards less distance last year! The lead has changed over the last furlong on several occasions in recent years (more frequently the places, for example just last year Big Fella Thanks went from contender to legless after the last, finishing 40L 7th) and so it could definitely affect the result but I don't think that it would negate the methodology of the modelling. Again, one can't anticipate the effect (subjectively one might estimate that it could help maybe 2 or 3 runners get a trip they wouldn't otherwise get but statistically that's very difficult to model on a hypothetical basis) but it should, over time, feed through via marginal changes to the algorithm as the model is tweaked to reconcile unexpected placings up to 30L of the winner. In the short term, it probably does increase the model's margin of error but, for me, it's a long-term exercise.
i thought the race was shortened by half a furlong, with the start nearer to the 1st fence, further away from the crowds, in theory making it easier for the starter as horses will be less distracted.
Cheers PBS, you're absolutely right. Sorry folks, make that approx 100 yards shorter. Still would probably have meant Sunnyhillboy beating Neptune Collonges last year, possibly Whats Up Boys holding off Bindaree in 2002 and Clan Royal doing the same to Amberleigh House in 2004. Can't say that for sure of course but certainly 3 of last 11 winners did take the lead in the last 100 yards or so.
Peanuts, you say 'my stat-model is not a process of elimination' , however, my recollection is that was the way you came out with the Welsh GN winner ? - could be wrong. Or is the GN simply a diff model ?
Peanuts, you say 'my stat-model is not a process of elimination' , however, my recollection is that was the way you came out with the Welsh GN winner ? - could be wrong. Or is the GN simply a diff model ?
You're right Chief, "my Model" exists only for the Grand National, which has unique characteristics. Re the Welsh and Irish Nats, there are some basic elimination rules which seem to work pretty well - helped me pick Bluesea Cracker for the Irish in 2010 and The Dude at Chepstow. Much cruder exercise though with much lower conviction as regards the selections. Interestingly, the age range preferred for both the Welsh & Irish Nats is 6~9. For the GN the sweetest age range is 9~10 and forget about 7 year-olds (6 year-olds not allowed to run any more in the GN of course) - reflects the fact that steeplechasers get further as they get older (even 8 y-os find the exceptional stamina-demands of a soft ground GN very difficult).
As expected, Tidal Bay, Outlaw Pete and Organisedconfusion scratched at yesterday's forfeit stage. They are joined by Call The Police, Pandorama, Little Josh, Lambro, Mister Hyde, Cross Appeal, Shakervilz and Fabalu. 11 out, 73 remain.
Comments
(I never bet e/w in the National - too many runners. I'd rather back another fancy than go e/w)
Anything e/w wait until the final decs are in and the bookies are offering as many as 5 or even 6 places
Prices don't drop for all runners, especially on the Exchanges. Dont Push It drifted way out to 90 or so on Betfair before the market got wind that McCoy might ride him (from memory about 2 weeks before the race) and Mon Mome's price was never longer than at the Off in 09 but if you fancy a horse that is a definite intended runner and is liable to attract interest (eg those that Walsh and McCoy may opt to ride - Dont Push It went off 10/1 Jt Fav!) then ante-post makes sense. For e/w betting I would wait until at least 5 places are offered.
Betting strategy is a personal matter but, for my GN system, I always back the 4 or 5 selections e/w with the bookies (eventually) and use the Exchange Win market as a trading/positioning vehicle. If you've got say 4 running for you, it's not inconceivable that 3 may pay off (blimey it happpened for my system in 2008 and 2010) and I'd do some serious damage to the TV if I'd backed them all to win and they came in 2,3,4.
Weights will go up at least 2lbs and Albertas Run presumably not a certain participant. Puts the cat among the pigeons. Wouldn't hurt T43 unduly as regards stat-profile, in fact could assist in pushing others further above 11.00.
All the runners will carry at least 2lbs more weight (maybe more if Albertas Run, the new topweight, doesn't run) but T43 is still good.
At least I hope there's 2 chances - an anxious wait for confirmation that T43's OK after his 43L 10th. Jumped and travelled well enough for the first 3m but maybe the notorious "Haydock Heavy" did for him in the last 1/2 mile. As you'd expect, Scholfield wasn't particularly hard on him once the chance to challenge went.
On paper not unduly concerned about the result. Mon Mome only managed a 42L 7th in the same race in 2009 before romping home at Aintree (with an even worse run in the Midlands National in between). Further back Earth Summit won the GN in 1998 after a 46L 5th in the Haydock GN Trial............maybe a good omen ;o)
For example, both Black Apalachi and Oscar Time ran well in the BJ before coming home runners-up in their same season GNs.
Notably, Black Apalachi also had a Becher Chase win to his name and had run very well in the prior GN and Oscar Time had been a close runner-up on heavy in the Irish National a year earlier.
Only Snowy Morning (3rd in both races in 2008) ticked neither stamina nor GN fence boxes prior to his GN placing.
So, what about this year's race, in which all 6 runners had GN entries?
Roi Du Mee won the BJ yesterday but has pretty dismal GN stats - in 22 chases he's never won a proper contest at any trip on decent ground and his only attempt at beyond 3m1f was a very remote 8th in last year's Irish National on Good-to-Soft. He has to carry at least 11.03 if he lines up at Aintree on 6th April and my model gives him little chance of being in the shake-up.
Price Du Beauchene ran another nice race and came in a close 2nd yesterday. While he has done little wrong since his move to Ireland and has obvious staying credentials in his family, he's never been tested beyind 3m1f and ticks very few GN boxes. He will have to carry at least 11.05 around Aintree (if Albertas Run comes out it will be at least 11.07). Don't Push It and Neptune Collonges of course carried 11.05 and 11.06 to win recent GNs but both had quality form at 3.5miles+. Like PdB, Snowy Morning had no form beyond 3m1f but at least his OR for the GN was 10lbs below his 155 mark when placed in the BJ. PdB has no such advantage.
Seabass did his 2013 GN stats no harm with his close 3rd yesterday and is a strong contender to be in the shake-up once again. However, his GN stats remain inferior to both Teaforthree and Cappa Bleu on any going but particularly if it were to come up soft. According to my model, a strong run but a place is his best prospect, barring others' misfortune.
Neither Rare Bob nor Bob Lingo should trouble the Aintree judges and, sadly, on stats aswell as form, Oscar Time is not the horse he was when a gallant 2nd to Ballabriggs 2 years ago.
This side of the Irish Sea, the Racing Plus Chase at Kempton was most notable for the truly awful jumping of Wyck Hill, apparently caused by a cut to a hind leg. He would need to do something special at Cheltenham to put his stats anywhere near giving him a chance to contend at Aintree. Otherwise, despite creditable efforts, there was little of relevance to the GN prospects of Quinz or Roberto Goldback, which are not great in either case.
It was a shame that the Eider Chase was snowed off. On the stats, a win or near-thing would have propelled 150/1 Tarquinius into the reckoning if the GN going were to come up soft~heavy. While he has entries at the Festival, it's possible Elliott may route him to Uttoxeter for the Midlands National in 3 weeks time; but he must win or go close on soft or softer there to give the required boost to his GN [soft~heavy] stats. In any event, there remains another funky outsider for Good-to-Soft or softer going at the GN - more anon.
First forfeit stage on Tuesday.
:-(
I take it that your ruling out due to distance? Five wins and a second in last 6 races ain't bad.
THE MODEL
A word of explanation about my model may be helpful. It has 2 important distinguishing features compared to most stat- or trend-based GN analysis.
Firstly, with one exception (any 7 year-old), my stat-model is not a process of elimination. It doesn't say, for example, no top-weight has won for 39 years so we’ll eliminate all top-weights. Rather, considering a variety of criteria, it applies penalties and credits to each runner based upon the degree of "mis-fit" or "strong-fit" of their career stats to the corresponding pre-GN stats of successful runners in all GNs since 1988 (excl the Red Marauder GN). So, being top-weight incurs a penalty but it is possible (difficult but definitely possible) that credits achieved for other criteria more than compensate for that penalty and give a top-weight a winning profile.
Secondly, the “successful runners” we’re looking to as statistical yardsticks are not just the GN winners but all horses placed within 4 lengths of the winners since, but for the vagaries of fortune, they could easily have won (4L is nothing over a 4.5 mile steeplechase with a big field). As an aside, in fact, the model takes account of all horses finishing within 30L of the winners, since their strengths and weaknesses and composite stat-scores (decreasing with distance from the winner) all help to inform the model’s algorithm. When the model is tweaked (after every GN) the aim is to reconcile unexpected results but also to ensure robustness. So, all tweaks to the algorithm are back-tested against all GNs since 1988 to confirm consistency of stat-scores vs distances for all horses winning and within 30L of the winners (of course allowances are made subjectively for meaningful mishaps during running).
The 2 distinguishing features of the model matter because they are the reason that it has (more or less) been able to cope with evolving trends in the race. There are two particular examples where the model has identified a GN winner that was eliminated by many trend-followers:
1. Prior to 2009, many routinely eliminated all French-bred runners because no French-bred had ever won the GN. But that overlooked the fact that 4 French-breds had gone very close indeed to winning in recent years: Encore Un Peu in 1996, Mely Moss in 2000, Clan Royal in 2004 and Royal Auclair in 2005 all finished 2nd, the first two by as little as 1.25 lengths. Then Mon Mome came in at 100/1 – lovely jubbly.
2. Prior to 2010, many routinely eliminated all runners carrying more than 11.01, since no GN winner had carried more for over 25 years. However, that overlooked the fact that 5 horses carrying 11.06+ had been GN runner-up since 1998 (two were top-weights). Increasing compression of the handicap had also led to the first 4 home in 2009 carrying 11.00+. Then Don’t Push It came in carrying 11.05 (followed of course by the French-bred Neptunes Collonges carrying 11.06 in 2011 - alas on this occasion the model preferred Sunnyhillboy!).
ROI DU MEE
You’re spot on about consistency of form this season and this is one of the criteria used by the model, for which he scores the maximum credit. He gets an additional modest credit for winning the Bobbyjo.
However, he has several negatives. He is modestly penalised for carrying 11.03 (that will get worse if the weights rise of course) and, again, for never having won a chase at a major left-handed track.
He is more significantly penalised because in 22 chases he has never won a proper contest on ground officially better than soft and never won any race (chase or hurdle) in a time equivalent (according to my model’s calculations) to Good-to-Soft or better. In 5 chases on (officially) Good or Yielding he has never been placed. That’s a big problem if the GN is run on anything better than Good-To-Soft because he will (according to my model) struggle to go the pace.
That penalty would be partially mitigated if he were to go close on decent going in the 2.5m Ryanair at the Festival, for which he has an entry, but it would need to be a 3m chase to remove it altogether.
One might say, he’ll be suited by soft going at Aintree as he’s done all his winning on Soft or Hvy. But in order to score as GN Soft Ground contender, the model also requires as a yardstick of the required stamina that a runner has won or performed creditably over at least 3.5 miles. The winners of all 7 GNs run on Good-to-Soft, Soft or Heavy ground (officially described or time-adjusted and excluding the Red Marauder GN) since 1988 ticked this box and only 1 horse finishing within 10L of the winners failed to do so (Garrison Savannah which had won the Gold Cup 3 weeks before and carried 13lbs less in the GN). Unfortunately RdM has done all his winning at no further than 3m1f. His only attempt at further was on officially Good-to-Soft in the Irish National (the winning time suggests that it was softer which should have suited) in which he was always behind and finished a remote 8th.
Again, that could change if Elliott gives him a run in the 4.25m Midlands National in 3 weeks time, particularly if it were to be on soft or softer.
However, there’s another negative stat for him on Soft~Heavy because he is an 8 y-o. In the 5 GNs since 1988 (again excl the Red Marauder GN) which were run on officially soft or heavy, or whose time suggested that the ground was such, no horse younger than 9 has made the frame. Indeed, of the 27 horses younger than 9 to run in these 5 GNs, only 4 finished (14.8%) [none within 30L of winner] compared with an average finishing rate of 22.8% for these races.
Having said all that, he’ll probably bolt up!
Apologies for the length of this – hope it hasn’t sent you to sleep.
In any event, you can only model the stats of actual races. While my system does seek to accomodate evolving trends it can only do so based upon actual results, with often the biggest clues to new factors coming by means of the placed horses.
The other change this year is to reduce the distance by a furlong (200 yards). May not sound much but I could have done with even 2 yards less distance last year!
The lead has changed over the last furlong on several occasions in recent years (more frequently the places, for example just last year Big Fella Thanks went from contender to legless after the last, finishing 40L 7th) and so it could definitely affect the result but I don't think that it would negate the methodology of the modelling. Again, one can't anticipate the effect (subjectively one might estimate that it could help maybe 2 or 3 runners get a trip they wouldn't otherwise get but statistically that's very difficult to model on a hypothetical basis) but it should, over time, feed through via marginal changes to the algorithm as the model is tweaked to reconcile unexpected placings up to 30L of the winner. In the short term, it probably does increase the model's margin of error but, for me, it's a long-term exercise.
Still would probably have meant Sunnyhillboy beating Neptune Collonges last year, possibly Whats Up Boys holding off Bindaree in 2002 and Clan Royal doing the same to Amberleigh House in 2004. Can't say that for sure of course but certainly 3 of last 11 winners did take the lead in the last 100 yards or so.
Or is the GN simply a diff model ?
Re the Welsh and Irish Nats, there are some basic elimination rules which seem to work pretty well - helped me pick Bluesea Cracker for the Irish in 2010 and The Dude at Chepstow. Much cruder exercise though with much lower conviction as regards the selections.
Interestingly, the age range preferred for both the Welsh & Irish Nats is 6~9. For the GN the sweetest age range is 9~10 and forget about 7 year-olds (6 year-olds not allowed to run any more in the GN of course) - reflects the fact that steeplechasers get further as they get older (even 8 y-os find the exceptional stamina-demands of a soft ground GN very difficult).
No pressure then.