IRISH BLOODSTOCK FORUMS :: View topic

Mr Jones»Wed Feb 23, 2022 17:04 pm
I think he only had 24 foals in his first crop - to run this year. Someone please correct me if I am wrong. I dont know any of them and I would be interested to know how many mares he covered in his second and third year at stud. Reply

Reference Point»Thu Feb 24, 2022 8:37 am
Yes, he didn't have many. They looked ok as foals. I can't seem to spot any in training. Reply

Bono 2U»Thu Feb 24, 2022 17:38 pm
He started off at £5k and is now down to £3k before his first runners. He may well get a few decent runners but imo his commercial prospects are dim. Reply

Reference Point»Fri Feb 25, 2022 9:28 am
If he gets a few decent runners this year his commercial prospects wouldn't be that dim. Reply

Squid»Fri Feb 25, 2022 11:41 am
Not sure about that Ref.if he gets a decent runner then most likely the dam will get the credit! Reply

Reference Point»Fri Feb 25, 2022 16:34 pm
I've never known that happen before (although it should happen) and not sure why it should be with him or any other stallion tbh. Anyway I'm keen to know if there are any in training as I can't seem to see any. Reply

TOPOFTHEHILL»Wed Mar 02, 2022 16:43 pm
RP for what its worth and at this time it may be worth very little, I have been test driving some new software which looks at deeper line breeding clusters up to 9 generations. Ragasinghe as a sire comes out with a very moderate profile and his cross breeding analysis is not so much better If you look at his 9 generation pedigree it would suggest you need a mare with a highish number of sons of Serenissima and Almahmoud, and daughter crosses of Pharos and Hyperion. Good luck with that. All in all it looks like an unlikely pedigree to get a really highclass runner but one thing this new software sort of implies if I am reading it correctly (and I may not be) is that you don't necessarily need a top class mare to breed a better than standard racehorse. Of course if you have a top class mare it might well be a lot easier but its not an absolute. The great thing about this new data is that it has substantial data profiles for really bad horses and they really do have quite different profiles to Elite runners. Its very interesting to get your hands on development software my guess is this is a longish way off the market or may well be released in a limited format which will be enhanced later. Reply

Reference Point»Wed Mar 02, 2022 19:55 pm
Many thanks TOTH. Hope you are well. I don't really have any interest in using him, I don;t have a mare for him even if I did! I quite like following 2yo's and I saw some nice foals of his and was looking to see if there were any early types. He was very precocious himself and the foals I saw were well stamped. I like the sound of your new software. It sounds like it may have some data on my horses Reply

TOPOFTHEHILL»Wed Mar 02, 2022 20:47 pm
RP I am sure that you are wrong these are genuinely slow horses . A lot of horses that just don't win races are IMO just unlucky they come from the vast majority of the population that is classified as Standard in this new Application and there is little difference in real ability or merit between them. If they find them selves in a race with 10 similarly talented horses its just the little differences that make the winner and I really believe the margins are very small. The difference to the Elite horse is however huge. The new data may just highlight a way to get a mare from a Standard to a higher level of ability. Its incredibly interesting to look at the same old problems through a completely different metric. Reply

Reference Point»Thu Mar 03, 2022 11:58 am
That sounds interesting TOTH. If you are looking back that far are there particular families/mares that are appearing frequently? I would assume Lady Josephine, Goody TwoShoes etc would be prevalent. Any surprising ones? Reply

TOPOFTHEHILL»Thu Mar 03, 2022 14:47 pm
RP as you might expect its far from simple! Firstly the thing that hits you is there is a difference in profile between colts and fillies (at Elite level). There have been some stand out features for top class runners but I am a bit reluctant to pass it on just yet as I need to verify the data. I had a look at the differences between successful sires and duds and again there appears to be some very obvious differences but then you get one that is either successful and doesn't obviously fit into the pattern or one that was useless and that doesn't fit into that pattern. So maybe I need to look again at what makes up a good sire or a bad one and see if they really do fit but not in the obvious way. Or it might be that the whole thing is no good, I am just not sure and need to do a lot more work. One thing I am certain of is that its going to have sod all commercial value unless we can start to identify much less complex patterns on a consistent level. there will be so few people who understand pedigrees that far back, its hard for me and I spent years researching longhand the old families and stallions. The footware clan is not as prominent as I might have expected but again I just may not have been looking at the right horses. I always thought they would be universally important in all elite runners. Sorry to be so vague but it is a completely new way for me to look at horses pedigrees and its a bit of a leap into the unknown. The developers told me almost nothing but the very basics and are relying on me to work it out, I am not sure if they want me to confirm what they have found or if they haven't really worked it out either and are looking for people like me to do their work for them. Reply

amadán»Mon Feb 28, 2022 15:21 pm
Return Of Mares 2019 ... 2019 coverings of registered mares 30, unregistered mares 3 Return Of Mares 2020 ... 2020 coverings of registered mares 22, unregistered mares 3 Return Of Mares 2021 ... 2021 coverings of registered mares 30, unregistered mares 6 Reply

amadán»Fri Mar 04, 2022 14:11 pm
I see TOPOFTHEHILL is using software that analyses deep into pedigrees. ----------------------- In Jan/Feb 2017 I wrote software that analysed inbreeding in six generation pedigrees, and compared 159k horses to their inbreeding. It proved certain types of inbreeding are positive, others negative. This work was proved statistically. There were a few errors in the programs, and a logic error in identifying inbreeding (known to few). Of course there were good horses that the program was unable to explain. ------------------------------ In the first five months of 2020 (lockdown) I wrote a ten generation inbreeding analysis program. It gave good information on great horses, but I have yet to produce a one number result for each horse. A 10 gen program is 16 times the size of a 6 gen program. -------------------- Improving speed is a major objective. Why? I do not analyse one horse. I analyse millions - the foals that all the mares at the sales could produce with all the stallions at stud (in Europe). I have a 12 gen program but have yet to find a use for it. My focus on improving speed reduced the time to generate a 10 gen pedigree file from about 0.80 seconds to 0.35 seconds then to only 0.03 seconds a pedigree. Speed improvements are vital. I am thinking of not doing the analysis on each pedigree as I move from horse to horse. Avoiding analysis will save time. Instead I will generate data from each pedigree, and do all the analysis afterwards. ----------------------------- I gave spreadsheets of my Jan/Fen 2017 analysis and my 2020 analysis to about half a dozen people, pedigree authors and pedigree analysis websites. One author told me to publish. ------------------------- This week I finished the collection of more ratings: 46,000 racing careers (generated about 16k best ratings) +8k +3k +2k +2k, = about another 33,000 ratings. -------------------------------------------- I have snippets of analysis that I will build into the 2020 program: the degree of close relationships between 400k horses identified (full; 7/8; 3/4; and other siblings - this can be complex with many subtleties); the average rating by position of inbreeding (only for the nearest inbreeding group) - very informative/educational; the explanation for some average runner sires unexpectantly becoming top sires. --------------- Yesterday I saw for the first time my colt foal born on 13th February. The mare is going from Ireland to France soon to visit a 2k sire (newish) that few use. At the stud many of the mares were due to foal to 50k to 100k sires and I wondered if any of the foals born will live up to those fees. They will probably sell well at the yearling sales. Reply

TOPOFTHEHILL»Fri Mar 04, 2022 15:53 pm
Amadan I am not using this software in a real sense, just trying to find if it is really useful or not and frankly I am still not that sure. It involves a level of complexity that would defeat a big number of breeders, and at very best will only be understood and appreciated by a very few, which is why I think it may be commercially unattractive. Like you I always keep an eye on what is going on further back in a pedigree but I have by and large adopted a pragmatic approach. If there are no indicators to potential success close up I don't waste too much time looking deeper. Some stallions are just not very good and some mares never produce a worthwhile runner. Its when you look deeply into a pedigree and find a sire for example that should be a solid sire because all its pedigree background is similar or at least aligned to other highly successful sires, turns out to be a complete dud. Then you find one that should be a dud who is knocking out winners by the score. Its very frustrating. I am sure that if I went further back I might find some clues but at the end of the day the dud one is still dud and the other one still a good solid sire. I think there used to be credible reasons why some sires didn't succeed due to getting the wrong mares but stallions today cover 4 or 5 times as many from all sorts of pedigree backgrounds its unlikely that they wouldn't find at least a few mares that would 'work'. Obviously the developers of this new software project would like it to be a commercial prospect but unless there is a market of pedigree people who are interested enough to spend countless hours researching, testing, or otherwise verifying the data its not going to happen. Maybe someone will find a simple pattern or three that is easily understood that will cover a big number of mares and a selection of stallions. I haven't done so yet but I am plodding on in hope. Reply

amadán»Fri Mar 04, 2022 16:41 pm
It is difficult to identify what makes a horse useful. There is always a pedigree explanation after the horse proves itself on the racecourse, but providing a pedigree explanation before the horse races, or before the foal is born is more difficult. ------------------- One thing I notice with a few good horses is a few ancestors low down in the dam match well with the 4th dam of the sire of the horse. An example is Sole Power, who has only one inbreeding in six generations, 5x6 Northern Dancer producing two sons (Nureyev; Lyphard who actually are two of his best sons.) But match the 4th dam of his sire Kyllachy (Mandragora (f)) with the 4th dam of his dam Demerger (Treaty (f)) in a mini pedigree between those two, and you get the reason for Sole Power's ability. ----------------------- The chief component of this mini-pedigree between Mandragora (f) and Treaty (f) is two groups of full siblings. The most important is the full sibling sister Imagery (f) (Gainsborough - Sun Worship) producing a son, and her full sibling brother Solario (m) (Gainsborough - Sun Worship) producing a daughter. The second full sibling group in this mini pedigree is full sibling brothers Pharos (m) and Fairway (m), but also Pladda (m) who is very closely related to the two but in an opposite way (son of Son-In-Law and daughter of Phalaris and daughter of Bayardo ----- whereas they are daughter of / son of / son of. ------- All these full siblings are in the 8th and 9th generation of Sole Power, and a quick look at his pedigree will miss the connections. -------- I have noticed this in another great horse - 4th dam of the sire connecting strongly with minor / obscure horses in the dam. ---------- Could anyone predict Sole Power's ability from his dam, Demerger, who was unraced, and who produced horses rated 103, 94, 67, 66, 40 and Sole Power rated 121? She produced ten foals, only one by Kyllachy, her 3rd foal Sole Power. She was sent to mate with four other sires after Sole Power had won his first Group 1 in 2010. Her last four foals from those other sires were either unraced or won nil (one finished 8/8 and 8/8 in his two races to a 30k cover (a "better" sire) when Kyllachy's fee was 15k). Reply

TOPOFTHEHILL»Fri Mar 04, 2022 16:06 pm
Amadan good luck with your foal. Of course the major problem for commercial breeders is to get financially sustainable and in order to do that they need to use commercial stallions that get top end sales results. It has almost nothing to do with racing but its how it is. I have always been fascinated by top end owners who would never use a stallion that might suit their mare if it was a cheap stallion, and I have some sympathy for that position. In a way its sort of easier if you have a moderate mare to build a strategy around a less commercial stallion. Also you are going to stand no chance of getting the moderate mare into a really top stallion even if all the analyticals suggest that it should or might work. I ran an old mare of mine through this new software and she had a pedigree that was comparable to the very best stallions even though she was actually a very moderate racehorse. I was surprised at first but then thought the reason I bought her was her back pedigree was very well connected. It was a very old fashioned uncommercial pedigree but traced to Classic runners and winners not to far back. She bred me a couple of quite good horses but if I had this data then I might have been inclined to breed her up a bit. Who knows what I might have achieved. Reply

Gone West»Fri Mar 04, 2022 17:41 pm
This is a fascinating thread and I applaud all hard research by you guys. Can I throw a question to you. Can you explain how - and it happens often - how two full-brothers trained by the same trainer etc. can have one brilliant and the other useless? Reply

amadán»Sat Mar 05, 2022 10:56 am
This is a point often made. One full sibling is compared to another. It is as if one example of difference between full siblings is enough to disprove all pedigree theory. ------------------- J B Robertson, a vet who switched to writing about pedigree, compared dozens of full siblings in an article in the book, The Lonsdale Library, Horseracing. ------------------- He found many useless full siblings of great horses. His work was excellent in that he did not just compare those that could be easily found i.e. where both had racing records, but traced all the full siblings in the stud books, many unraced. Mordaunt Milner in his book Thoroughbred Breeding (1987) mentions that research and arrives at much the same conclusion as I did. There are a great number of excellent full siblings, and many poor full siblings. But the chance of producing a top runner is much greater if the mating that produced the top runner is repeated. J B Robertson's first example was two fillies who both won the English Oaks. --------------------------- I answered one person who told me Sadler's Wells was a much better runner that his full sibling brother Fairy King. My reply was Fairy King fractured a sesamoid in his first race and was retired. ---------------------- This week I watched all the 2006 Irish Group and Listed races. They were on a DVD from Irish Thoroughbred Marketing. I noticed a very poor performance in a three runner Group 3. Tiger Dance finished third, 18 1/2 lengths behind the two who battled out the finish separated by a neck. I looked up Tiger Dance's race record and saw in his previous race he finished 4th of 4 in a Group 3 beaten 9 lengths. For some reason he was rated 100 for a 6th place in a Group 2 beaten 8 3/4 lengths. The real surprise to me is he was a full sibling of Giant's Causeway (Storm Cat - Mariah's Storm). ----------------------- What about all the full siblings produced by Storm Cat - Mariah's Storm? They were Freud (rated 114); Giant's Causeway (132); Pearling (76, two races); Roar Of The Tiger (100); Tiger Dance (100); Tumblebrutus (104); You'resothrilling (114). It might be of interest to note that Pearling (76) produced Decorated Knight (rated 121, won three Group 1s). ------------------ I would much prefer to repeat a mating that worked than try something different (what I call sire bingo). Reply

TOPOFTHEHILL»Sun Mar 06, 2022 11:09 am
I would agree with most of that its like we always say where is the best place to look for Gold? Where someone else has already found it! Reply

amadán»Sat Mar 05, 2022 16:05 pm
I wasn't too satisfied with my reply as I only used a few full siblings to show how alike their rating were. --------------------- I have 1,766 horses by Sadler's Wells (1981) (m) in my ratings file (not all with a rating). Of these 1,134 are full siblings in 416 groups. I spent a bit of time looking up about 150 of the 1,134 that had no rating. Many were NR (non-runner), a few raced in GB, FR, USA without the rating agencies there recording a rating (i.e. many ran very poorly). ------------------------- The visual impression I got when filling in the gaps was how alike full sibling were in running ability. ------------------------------------------ I will list a full sibling few groups to show that [85;usa] [111;13;63] [101;104] [85;nr;96;80;96] [74;113] [80;99] [74;89] [85;119] [121;108] [104;95] [66;30] [fr;89;98] [105;73] [108;58] [62;nr] [jap;114;114;57] [92;78;110;75] [106;78] [nr;78] [102;114;105] [92;nr] [68;84] [104;121;114] [115;88] [109;96] [109;95] [100;nr;50] [56;109] [ity;70;99] [76;69] [fr;83;72] [76;103;fr;96] [95;97] [nr;72;120;nr;65] [nr;101;122] ------------------------ I did not separate the fillies from the colts. My guess is the fillies probably rate lower. Reply

amadán»Sat Mar 05, 2022 16:42 pm
To get a better idea I decided to look at the bigger groups of full siblings (six foals or more) sired by Sadler's Wells. I will not give the names of the dams that produced these full siblings with Sadler's Wells. ---------------------- [nr;72;120;nr;65;can] - [nr;82;117;58;94;116;90] - [116;109;110;nr;80;gb] - [124;80;113;100;nr;nr;nr;nr] - [85;73;110;100;82;70] - [119;124;99;124;24;112;nr;86;nh] - [87;102;75;117;127;116] - [nr;79;fr;111;85;113;79] [64;45;nr;107;116;75;15] - [114;91;97;nr;106;82;119;nr;119] - [113;97;97;99;132;80;93] - [95;83;84;88;98;129] - [104;113;96;87;76;76;67] - [72;107;75;65;128;92] - [86;79;84;92;80;103;113;nr] - [86;126;nr;nr;nr;nr] ----------------------- nr = a named non-runner / gb; can; fr; ran in those countries without getting a rating / nh = ran in national hunt only. ---------------------- In both cases where a dam produced four non-runners it was 3 fillies and 1 colt. Overall in the examples above it was 3 colt non-runners, and 15 filly non-runners. ----------------------- I did not use averages as imo there is no such thing as averages in small samples. What I might do is investigate why a few fillies produced better colts, and other produced colts and fillies of equal ability. Reply

Reference Point»Fri Mar 04, 2022 18:14 pm
Yes, those are good points TOTH. What about some analysis around mares rated 80 or less to see if improving those has any patterns. Good luck with the foal Amadan. Reply

TOPOFTHEHILL»Sun Mar 06, 2022 11:22 am
The difficulty here RP is the rating system for the vast majority of horses in the middle of the rankings i.e. those not too bad but a bit off being smart horses. There is really nothing much between them it is certainly not the number of lbs difference they are in the handicap which is why its so damned hard to find a winner in a big field handicap. If you try to make sense of the ratings you will fail, OK they may be useful, especially at the top end of the handicap but 75 - 105 its really anyone's guess. I think in a way the handicap system is as good as it can be but it is analogue in the digital age. There are plenty of punters who struggle with it in its present form but in my opinion, and this is only my opinion, the ratings need to be graduated so that the differences between horses in that huge average group are calibrated in 100ths of a unit while at the top and bottom the full unit may express the difference. In the intermediate levels of ability something in between seamlessly expresses true form. I don't know how you would do it but we put people on the moon. We have created weapons of Mass Destruction so complex that it is beyond the comprehension of many. We have developed solutions for many much more complex problems. Why not. In the meantime RP I will try to do something for you but remember my opening post it would appear that you don't need a top class mare to produce an elite runner but its a whole lot easier if you have one. Reply

Reference Point»Mon Mar 07, 2022 12:40 pm
Thanks for reply TOTH. Yes the ratings are very often a ballpark figure of a horses ability at best. I would assume ability does matter if you could measure it accurately? It will be interesting to see if you research will produce some correlation between good horses and something other than the mares ability. I tried to buy a mare in December but looking for something with a decent level of form (close to 90) with some commercial appeal was tough. It felt like more breeders were looking for mares with decent ability than good black type pages. In the past I have always thought it was the other way around, but of course I wasn't trying to buy one then. Reply

amadán»Mon Mar 07, 2022 13:22 pm
I share TOPOFTHEHILL's uncertainty about ratings. I now have 167,908 ratings on file. They probably include a disproportionate number of high rated horses because high ratings are easier to find. ------------------- This is the split in 10 point bands e.g. 0s is from 0-9; 10s is from 10-19. ------------------- 0s 0.2%; 10s 0.5%; 10s 0.9%; 30s 2.4%; 40s 6.8%; 50s 11.7%; 60s 15.2%; 70s 17.2%; 80s 14.6%; 90s 12.0%; 100 10.2%; 110s 5.4%; 120s 1.8%; 103+ 1.1%. ------------------------- The almost non-existence of horses rated under 40 probably is due to the reluctance to give a horse a miserable rating. Better to not give it any rating. --------------------------------- Make up your own mind about this: rated 79 (2579 horses); 80 (3145); 81 (2262); ----- rated 89 (2007 horses); 90 (2444); 91 (1880) ----- rated 98 (2038); 99 (1857); 100 (2076); 101 (1857); 102 (1880) [1857 horses for 99 and 101 is not a typo, just a coincidence] --------- see how the number of horses rated 80 and rated 90 and rated 100 are greater that those either side of those ratings. It looks to me like handicappers was doing favours rating a few horses at the magical numbers of 80, 90, and 100. Reply

Reference Point»Tue Mar 08, 2022 10:22 am
Hi Amadan. I think that's to do with grades. If your horse wins the handicapper has to work out how good the win was and how good tour horse is. Say you win a 0-70 off 66 the handicapper has to penalise you but if you are a known quantity (to him) and clearly can't win in a grade higher he will give you 70 (top weight in the grade)and then you start dropping you back down to an eventually winning mark. Reply