2017 NMP Service - post season statistical breakdown

E-mail first sent to NMP members on 21st November 2017 and discusses the areas where the service under and overperformed in 2017....

Overall results 

The 2017 service, which began on 9th November 2016, ends with a final profit of +4.33 pts from 1115.48 pts staked (+0 % ROI - break even to worst prices or a small profit to anyone (i.e. most) who took prices and achieved better than that in reality).

Seasonality - a game of two halves 

Service profit from Nov '16 - May '17 

537.50 pts staked 

-35.09 pt loss 

Service profit from June '17 to Nov '17 

577.98 pts staked 

+39.42 pts profit 

Analysis 

I tried to emulate the memorable Cheltenham of 2016, when we made +25 pts profit from 50 pts staked (to lowest prices), but this time we didn't make Cheltenham pay. 

The crown jewels in our ante post betting list - Unowhatimeanharry (16/1 into 10/11 fav) and Yanworth (12/1 into 2/1 fav), surprisingly, both lost. 

We also wasted stakes trying to get some of the Cheltenham bets onside in multiples before the Festival had began (an approach which worked brilliantly in 2016 - but didn't work so well this year).

Instead of starting the turf season over 50 pts up, as we did in 2016, this time we were -26 pts down. We therefore started this season on the back foot. 

The way the service sets up, we need to be landing the big Saturday winners if we are to enjoy a successful season. In the first six months of the 2017 service, we didn't have any bets which yielded a double figure profit*, and only eight winners in that period produced a profit in excess of 5 pts. 

We would expect the first six weeks of the season to be difficult but this year more than others, we just couldn't get going early on. By contrast, in the second half of the season 17 winners returned a profit which exceeded 5 pts and we had five of those which were double figure returns. 

*for this aspect of the analysis each bet was treated individually. So if we won two lots of 2.5 pts from two different bets (for example if backed in a 0.5 winning single and included in a seperate winning double) it wouldn't have exceeded 5 pts for that particular selection 

Just looking at the staking totals, I feel we overstaked in the first half of the year. We should be turning over higher stakes in the summer, when the quality of racing is higher, but a heavy involvement at Cheltenham - both during the week itself and leading up to it - has upped our turnover in the first half of the year. 

Moving ahead, as stated previously, we will be cutting the turnover over the winter and for the Cheltenham ante posts this year. More of our Festival bets will be placed during the week itself. I am also looking forward to not moving house next year! Things have certainly settled down from July onwards since we moved in and the decorating and prep work came to an end!

Performance of the singles vs multiples 

Singles 

971.38 pts staked 

+25.05 pts profit 

+3 % ROI 

Multiples 

144.10 pts staked 

-20.71 pts profit 

-14 % ROI 

Analysis 

Firstly, the singles have obviously underperformed. Bearing in mind the service had an +11 % edge at the beginning of the year, for them to produce only +3 % ROI this year is disappointing. 

I have subsequently looked into the performance of the singles further. As stated above, not enough double figure winners were landed in the bad run in the first half of the year. Things have picked up in the last three months, however, with Nakeeta (Ebor), Donjuan Triumphant (rearranged Ayr Gold Cup) Withhold (Cesarewitch) and Saunter (November Handicap) all going in.

In terms of the multiples, I've touched upon this issue many times. When looking at the -20 profit figure it is worth remembering most of the multiples involved settling at SP - so anytime a horse was backed, the P/L column bore the brunt.

The nature of multiples (big cumulative odds) also means longer losing runs and extreme P/L swings are more likely - they are more volatile bets. The P/L figure for the multiples also doesn't fairly factor in the number of near misses we've had. 

A couple of the 'nearlies' were.... 

Home of the Brave - traded at 2.24 in the Lennox Stakes but beaten by a winner who had hit 1000.0. Victory was worth over 20 pts 

Jumira Bridge - traded at 1.18 and was beaten by a neck by a winner who'd hit 20.0 (having gone off 3.00). Victory was worth over 20 pts. 

Orion's Bow - hit evens at Newcastle but endured a nightmare run. Victory was worth over 10 pts 

Growl hit 1.42 in running in the Stewards' Cup, before finishing fourth under soft handling. We were on for 40-50 pts if he had won. 

.......based on the in running odds the above horses had worked themselves into in their races we could have expected 2.5 winners, so to get none was very harsh (and there have been other close calls too). 

I'm now keeping separate P/L tallies for the singles and multiples each month so members can see how both type of bet is faring. Opinion remains polarised but hopefully members who want to back the multiples for that excitement of standing to win big still will do so, while those who want a smoother, steadier ride, can ignore them and stick with the singles. 

I will still send multiples in the 2018 service, but I don't intend on turning over quite so much on them next year. The danger, as ever, is that if one of the legs wins, you are relying on subsequent legs to get a return so, in that respect, you need a greater element of luck with these riskier bets. It makes sense to have the bulk of our money on the selections where our edge is more pronounced, i.e. the singles and the fate is less reliant on luck. 

2017 Profit breakdown by odds range (single bets only):

Odds of 10/3 or less: 

264.25 pts staked 

256 bets 

+41.45 pts profit (+16 % ROI) 

Actual winners 88 

Expected winners 79 

A/E 1.11 

Odds between 7/2 and 10/1: 

555.13 pts staked 

621 bets 

+2.13 pts profit (0 % ROI) 

Actual winners 98 

Expected winners 98 

A/E 1.00 

Odds of 11/1 and bigger 

152.75 pts staked 

198 bets

-18.53 pts loss (-0.12 % ROI) 

Actual winners 13 

Expected winners 14 

A/E 0.93

Analysis 

The shorter priced picks have performed very well this year and it is the picks at longer odds which have held us back. The point to note with the longer priced picks, however, is that the potential returns on such bets are greater, and that minus figure could have quite easily swung into a profit if we'd enjoyed a 'normal' (i.e +11 % ROI) year. 

The 'expected' figures, based on the advised odds of the selections, suggested we should have had at last 14 winners within the 11/1+ odds bracket. We actually managed 13, which is one less than even a service with no edge could quite feasibly manage. 

With an 11 % edge going into this year, I think we have probably been short changed of two, possibly three winners at odds of 11/1+ this year. Growl and Unowhatimeanharry or two who immediately spring to mind! 

I don't think the number of winners at odds of under 7/2 flatters us, as an 11 % improvement on the expected number of winners is 87, and we achieved 88, which is close enough. This particular odds bracket performed perfectly in line with realistic expectations. Even if you took the view we had one 'lucky' winner here, it would still show a generous profit if that extra winner was discounted. 

With regards to the 7/2 to 10/1 bets, which accounted for a good chunk of the turnover, we broke even, with the number of expected and actual winners leveling out at 98. We know the service has a +11 % edge long term, so an 11 % improvement on 98 expected winners is 109, i.e. we could normally expect 11 further winners in this mid odds range. 

There have been plenty of near misses form the placed selections who didn't go on to win, so it is easy to imagine these selections faring better if we'd had the rub of the green this year. 

2017 Profit breakdown by stake strength (singles only):

Top of page