There is an old thread on this, but we’ve been thinking a bit about changing things up next year and I was wondering if anyone had any new thoughts on what the ideal release time (of our outright odds, matchup odds, etc) would be.
I think a Tuesday release of all model predictions would be better for matchups. Right now we only technically post matchup odds and the h2h tool on Monday night, but because we’ve already posted the skill decomposition page it’s not hard to ballpark h2h odds before then. Obviously a Tuesday release would not be ideal for people who bet on the exchanges, or who are mainly focused on outright and place markets. Although I think a lot of the place markets are released after we post our model early Monday afternoon ET.
We can of course keep things as they are… it’s just annoying to see the usual suspects copying our numbers for their opening lines this morning.
My perspective may be different as I mostly bet in markets that aren’t directly in the scratch tools and the model gets released in the evening my time.
For PGA and Euro weeks, The biggest bookie have their main competition fully up and the secondary mostly up by model release. There is already movement on the odds well before the model release so there is missed opportunity as is.
The times that I’ve picked up significant value on matchups and 3Balls is on independent pricers that have had markets on more obscure players that didn’t have many other bookies running odds.
If it is pure hunt for value, maybe it is reducing the number of scratch tools info for a period but allow people to find it themselves. Only you will know if your scratch membership are hardcore enough to prefer doing the heavy lifting but get greater reward though.
I agree a compromise is maybe don’t put the odds screen up or the player skill decomp till Tuesday (or Wednesday) and just have subscribers use the custom matchup tool to manually punch in the early matchups.
The easier it is to use, the less valuable it is unfortunately. Kinda an existential problem when you’re trying to run a SaaS company lol.
You could also just raise the price a bunch and work more closely with a smaller group of customers. I think honestly I may rather pay $1k a month for DG than $30 as wild as that sounds - and the price would only go up if I knew the books were being screened out because you were interviewing customers or something.
I agree with @Obgeoff, there is already alot of movement in the markets before the odds are even released. As soon as B365 post their ‘Miss Cut’ bets, the good ones are eaten into straight away.
I think if the odds get posted too much later all the really crazy value will be gone by the time we get to see the models predictions. I know for a fact there are other groups out there with similar models, if not more accurate ones, so posting the odds later will just put us at a disadvantage Vs them.
I think the best compromise is to release the model early and refrain from updating the scratch tools until Tuesday/Wednesday (except for the Tournament Props)
That way we can simulate all the matchup and 3-ball odds manually and still look at the model predictions for the finishing odds, but books can’t update all the numbers within 10 seconds
Without the scratch tools the books are much slower.
The reality is the front-facing and betting-focused nature of the site means this will impact the market, there is no way around it at this point. You can’t screen customers or anything to prevent books from getting their hands on your numbers once they are out in any capacity. Releasing only the skills decomp doesn’t really accomplish much besides making it tougher for customers to find value since, as Matt mentioned, you can easily figure out H2Hs and outright fairs from those.
The only way to have minimal market impact is to wait until after BM/BetCris posts their H2Hs before releasing any information - from there on, they are the market-makers and DG release will not move numbers very much. The question from there, of course, is whether or not there is significant value remaining that late. Personally, I think the answer is still “yes” but a huge majority of DG’s predictive value is picked off at that point. In fact, since BM/Cris tends to release mid-to-late Tuesday, a vast majority of customers are likely not getting a look at the screen until Wednesday. If people are strictly using DG for its predictions, I believe they are going to be betting neutral EV, at best, in that scenario.
While DG isn’t a tout, per se, this is a classic tout issue. “If I’m good at predictions, how can I sell my information to people while there is still value, while also not indirectly selling to books?” The reality is you can’t. Obviously you can shut down all the predictions and have no impact, or you can wait until your model doesn’t have value anymore, or you deal with impact.
A couple of reflections. Bet365 basically puts out whatever their own models gives them. Regardless of other books odds. And regardless of sport. And BM/betcris have no problems exposing them selves quite a bit even way after scratch tools is out here. There’s examples of that every week.
In genereal though, I actually do think it matters if the scratch tools matchup is released or not. The biggest line changes happens after scratch tools release. So I guess I’m leaning on DanielSongs thought I guess.
I think things are working pretty good as it is though, but if anything I would much rather have tourney props tool up to date before outright/matchup. And a updated version would be much appreciated. Especiallyt being able to customize and combine nationalities since there’s always all kind of cross over markets like top scandinavian, GB+Ireland, Asian, south american, and so on.
Like things as they are. Yeah it sucks to miss T20 EV if you’re not attached to your computer and doing it immediately, but that would be the case no matter when you release (and probably missing even more +EV the later you publish).
In theory, there will always be value, right? As your tool improves, especially if there is progress in alt markets tools like @madmartigan5 is talking about, you can always find a number.
The advancement of golf betting is good and bad. It’s more popular, and more books carry more lines, but it’s made it tougher to gain an edge. But with good tools like this, there is always an edge somewhere.
I think a Tuesday morning release would work well. It won’t stop the few bookies copying that release late on Tuesday but it will mean that the UK books can’t copy straight away which is then far better as those prices probably act as a base for a good part of the US industry prices.
One slightly different way to look at it though would be to try and enforce a personal and an enterprise licence. Almost every software package or website has that pricing structure when private companies use it. I suspect it would be very hard to enforce but it is relatively obvious when someone is using datagolf prices. Some companies still pay 5 figures to a wholesale provider to get worse prices than what datagolf offer for $200. So it seems to make sense that other companies will just pay the $200 but I suspect you could make them pay a lot more.
Mainly an outright and place market bettor here so I agree that current timing works well and wouldn’t want it going any later as value would disintegrate over time. With the current release time most books have outright markets posted and there is usually decent value.
Would also be interested in paying more and having some sort of verification process, but understand that this is logistically very very hard and probably not the best for DG’s bottom line.