FWIW in a theoretical all Luke final, I'd put Littler at around a 2-1 favourite, although the numbers are all over the place. Mid sample appears 50/50, so you can only guess what Littler is doing in the other samples to make it balance at that overall projection.
Still not liking bad commentary. Still not afraid of double nine. Just a bit more subtle about things.
Saturday, 11 October 2025
Any semis value? Maybe
Looking at the 365 lines, they've got both Lukes priced in at 1/4, which is an 80% chance of winning. That seems a tad excessive. Noppert and Clayton are not bad players at all. I've got both of these games at nearer the 70/30 marker - Humphries a touch above 70%, Littler a touch below, maybe in a realistic world Clayton could be priced a bit shorter by default but the pricing is such to account for the Littler hype/public backing, which is not unreasonable. So would I be betting on a Luke/Luke final? It seems by far the most likely outcome, I think it happens clearly more than half the time, but I certainly wouldn't be wanting to bet on either of them to win their respective matches. Would I want to bet on either of Clayton or Noppert to pull the upset? This is a more pertinent question, and I think the choice of the two would clearly be Clayton, but I'm not sure that you'll actually get the odds to offer sufficient safety in what you're doing to actually be profitable.
Friday, 10 October 2025
GP quarters
Christ, what a turgid set of round two games those were, nothing to a decider and four whitewashes. Still, warms us up for the new expanded worlds I guess. Any value in the quarters now?
Noppert 34-66 Anderson
van Duijvenbode 45-55 Clayton
Littler 67-33 Price
Humphries 76-24 Menzies
So no, no value anywhere really.
Wednesday, 8 October 2025
GP R2
Should I even bother with this? Basically all the projections ended up bang on the bookies' lines. There were some exceptions, Dirk being the huge one, but outside of DvD it was very slim pickings, Cullen, Gilding and Rock maybe being fractionally undervalued in the market. Let's take a look anyway,
Menzies 30-70 Cross
Bunting 74-26 Noppert
Humphries 78-22 Ratajski
Cullen 26-74 Anderson
van Duijvenbode 70-30 Gurney
Price 43-57 Rock
Clayton 79-21 Woodhouse
Littler 80-20 de Decker
So, at a quick glance, it's pretty much Dirk and Rock again, although it should be noted for Dirk that the shortest form sample is pretty much bang in line with the market and it's only larger samples that push him to being as big a favourite as I suggest. Otherwise, it's mainly favourites being perhaps a tick or too on the long side, but nothing really exciting.
Monday, 6 October 2025
Grand Prix round 1 projections
I'm just going to post round one at this stage. Too many permutations for round two or later, no rush for those at this stage either. Two big caveats:
1 - THESE DO NOT FACTOR IN THE DOUBLE IN FORMAT. Frankly my data model is entirely setup for straight in double out. At this level of player, I don't think it's going to make an enormous amount of difference, everyone's likely to be somewhere in the one in three to mid 40% range on doubling in (it should be a bit higher than doubling out for obvious reasons), which I think is a tight enough range not to make any serious modifications. If you do have reliable data (and I don't think you have) and you are seeing a big difference in any game feel free to shift things by a percent or two. I really can't see it making much more of a difference than that.
2 - THESE DO NOT MODEL THE EXACT MATCH FORMAT EITHER. This is the only tournament with a first to two set format, and with first to four or more sets with no tiebreak. For round one, I'm going to pretend this is best of 15 legs (which is the longest the match can go), and then when we get deep, just use the projections as if there was a tie break. These will push up the better player SLIGHTLY, but you would think the better player will be in general a better doubler as well, so I think those more or less cancel out.
In any case, here we go. Use at your own risk.
Luke Humphries 68-32 Nathan Aspinall
Martin Schindler 59-41 Krzysztof Ratajski
Chris Dobey 69-31 Cameron Menzies
Rob Cross 53-47 Wessel Nijman
Stephen Bunting 71-29 Niko Springer
Danny Noppert 52-48 Jermaine Wattimena
James Wade 63-37 Joe Cullen
Gary Anderson 79-21 Raymond van Barneveld
Luke Littler 67-33 Gian van Veen
Peter Wright 30-70 Mike De Decker
Gerwyn Price 58-42 Ryan Searle
Josh Rock 81-19 Ryan Joyce
Michael van Gerwen 45-55 Dirk van Duijvenbode
Ross Smith 67-33 Daryl Gurney
Jonny Clayton 68-32 Andrew Gilding
Damon Heta 70-30 Luke Woodhouse
Sunday, 5 October 2025
The ultimate Harrington visit
This is very much a fun, low content post, those expecting high value Grand Prix insights can come back tomorrow. But long term readers of the blog will know we're not fans of Rod Harrington here at Tungsten Towers. Heck, short term readers will also know if they've read the FAQ. In any case, it's not actually a name that's come up recently, until someone on X (I forget who and CBA to look through my archives, but you know who you are) highlighted that someone, I want to say Dirk, was on a simple single-double out in a recent Euro Tour, but missed a big number twice in a row to not get a dart at an out, which was described as "the ultimate Harrington visit" or words to that effect.
But no, was my retort. That would be if those two missed big numbers then left double nine, which Dirk then hits. But can we reasonably create a visit which does this? I feel we need some caveats here.
Firstly, you must be aiming at a reasonable target to set up yout outshot of choice. and it must be a reasonable outshot. Some people, like Littler with his D15 shenanigans and previous to that (the now resurgent it must be said) Mensur with D14, have pushed the boundaries of what is reasonable. Let's just have a bit of common sense. If you are on 40 and claim "going 8 for D16 and I pushed it into D11 I'm now on D9 gg" you can fuck off, you're going for tops.
Secondly, your miss must be reasonable. You might pull one into single 7 going for big 20, but nobody good is. The point is that Harrington would have a meltdown, and he's not watching your games at the Dog and Duck, and neither am I. Let's limit stuff to missing by one bed and that's it.
Finally, I think we should try to be consistent, we shouldn't really switch from being a tops and tens player to a 16s and breakdown thereof player mid visit for the purposes of getting a funnier route.
So, can we do it? Well I've got three that I think come pretty close. Making up the podium, and I think I just about prefer the first one, we have:
Requiring 41 - big 20 going for single 1, treble 1 going for single 1, double 9
Requiring 49 - big 3 going for big 17, double 14 going for single, double 9
Neither of those really do it for me - in both cases, the second visit has at least hit the number we're going for, even if we've not hit the single we wanted. However, there is one route which I think is the winner for now.
Requiring 43 - big 8 going for 11, big 17 going for 3, double 9
I think that's about the best I can come up with, looking at all the options for a second dart to leave 18, nothing else in the lead up seems to have either a legitimate neighbouring miss, or simply have a legitimate alternate target which we must go for (good job saying that hitting any form of 6 is a possibility on 24 for example).
This works, however, it still doesn't really cut the mustard for me, and for one reason. And that's the second dart. Now anyone at any level will probably have needed 35 at some point, and have been told "look if you want to leave D16 at least aim for the left side of the 3 so if you miss you still have D8" which, even at the pro level where they're missing big numbers fairly rarely (and is not only the entire point of the meme, but something someone ran some numbers on on the PDC site fairly recently), they'd probably tend towards making any misses miss left.
So can we do better? That's your challenge. Maybe there's some shank into an adjacent treble route I've completely misses or similar. Go for it. And if you are wanting Grand Prix numbers, I'm close to done with getting the players in that event up to date after PC28-30, even if I'm not close to the rest of the field, so check back later today or maybe tomorrow. I'm off work then so I should have something (which will be brief because double in format) before the off.
Thursday, 2 October 2025
The Tungsten Analysis Hot 100
So this is something that I have teased on X I think. Here I've done something really simple, but really enlightening. Potentially. I've taken four of the metrics I used - the FRH ranking which has been in play for probably a decade now (scary to say that), which for newcomers is a modification of the main order of merit to primarily favour recency, then scoring averages for each of the last 100, 200 and 365 days, which are fairly comparable to the portions I use when making projections. The newest goes back towards just before the Matchplay, the middle one to just after (IIRC) the UK Open, the last year is self explanatory. All these have a "one leg played per day in the database" qualifier. So, here we go:
Some things to note. This is not in any way intended to be clickbaitey, although I may make an X post that is just that. It is what it is. Some people were unfortunate in missing the leg criteria by not much - Wright recently being the most egregious example, but Tricole and Greaves also didn't miss some by much (the latter not hitting the list at all as a result). Brooks being 5 recently is not a typo, he is that good. And for reference, these stats were taken BEFORE this week's Pro Tour games.
Subscribe to:
Posts (Atom)