Rento Saijo

Turns out scoring 14 goals against Colorado in a series only wins you one more game than scoring 5 goals.

My backup bracket (which obviously doesn't count because you can only officially submit 1 to the Challenge) looks fantastic. My main one is still really great, though.

I was a little afraid to post this beforehand, but I can prove that this is my official submission to the bracket challenge (and therefore valid); the model + epsilon goes 6/8 in the first round:

@AirOnBiggio It wasn't intentional per @OHaraSports 😂

An OT winner by Brett Kulak playing with the Nichushkin-Drury-Necas line… Bednar Blender Masterclass. #GoAvsGo

Martin Necas tells @TheHockeyNews that he wasn’t even supposed to be on the ice for the last shift. He noticed they didn’t have enough skaters and made the split-second decision to jump on, and he makes the game-winning assist. UNREAL #goavsgo @TheHockeyNews #mnwild

@DavidCBets @FutureCanes @Tttaussig This is almost exactly what I said, but in 2020.

@FutureCanes Necas is legit going to be the best player on a team somewhere. He could be a McDavid from Wish for someone. Our scheme holds him back. Just best for both sides to go separate ways imo. We haven’t used his potential.

I know in terms of roster construction moving forward that it’s *near* impossible to lock up all 3 of Guentzel, Jarvis and Nečas long term but….Man I think you have too. Nečas, for me, is Lindholm 2.0. He will go elsewhere and explode, and you will regret it.

@heatdaddy69420 @Tttaussig

Marty Necas is better than McDavid and I’m being serious. No disrespect to McDavid I love Conor McDavidson. Go watch in real life.

Necas slippin' n' slidin' 😂

For the "second shot," I will be taking the Ducks (I will stick with the Avs against the Wild).

@ivandemigoal 😭😭😭

It takes some real talent to make a bracket like this

@Dar56810307 @hischiersdevils Ducks over Kings is a huge difference; not "only" lol.

@hischiersdevils Way to go and my only difference is ducks over kings and sabers over Canadians

am i about to go 8/8 😭😭😭

@guitarmack5 @hischiersdevils They can prove it if it's their official submission to the Bracket Challenge. Not sure if they will.

@hischiersdevils The first time I can find you posting anything about the brackets is 5 days ago…. Absolutely bullshit

@MikeKellyNHL My model also had Flyers in 6! 3/3 so far as well :)

The model called Flyers in 6, and so it was. Penguins racked up offensive zone time while the Flyers capitalized with quick-strike offense. Philadelphia will face Carolina in Round 2. The model improves to 3-0.

@drivelinekyle @thsottiaux @sama Not sure if you're looking for just Codex or ChatGPT included, but if we just Codex, they have a pay-as-you-go Codex only seat for businesses, I believe.

@thsottiaux any plans to make Business Pro seats? $20/month per user is great but we have users on the Premium level for Claude and they need expanded limits if we migrate to OpenAI. didn't see an option there. @sama

Don't just reset Codex rate limits for fun, it costs money. Don't just reset Codex rate limits for fun, it costs money. ... but the vibes are good ... I have reset Codex rate limits for ALL paid plans to celebrate a good week and allow everyone to build more with GPT-5.5. Enjoy

@JBott32 @KellenDB @MikeBartner @Sportsnet The struggle with these are accuracy, as with any other similiar systems.

@JBott32 @KellenDB @MikeBartner @Sportsnet I believe they already do have an RFID chip inside these pucks. "The cameras detect infrared signals from the pucks up to 60 times per second and the players up to 15 times per second, generating millions of raw location data points," per NHL website. (1/2)

Did some AI puck-tracking on that controversial Oilers/Ducks call. As an Oilers fan it pains me to say it, but the math doesn't lie — it's a goal. By 3 pixels. @Sportsnet

@TheRealFriscoJo @LinkofSunshine See: Bayes' Theorem.

@LinkofSunshine Can someone smarter than me explain how it's not 99%. I mean, explain to my friend, my friend's a little slow and doesn't get it.

The most worrying poll I’ve ever seen is one when they asked doctors “1% of the population had disease X. The test has a 99% accuracy rate. If someone tests positive for disease X, what are the odds they’re have disease X” And like 95% of doctors said ~99%. This is their job.

@StatsByStas @mattdonders @AlexC_NJD @CJTDevil @yolo_pinyato I do wonder with Open AI's Images 2.0, you can generate some cool designs? Haven't played around with it too much, but I'd be curious.

@mattdonders @AlexC_NJD @CJTDevil @yolo_pinyato Making my own website also using Claude, basically converges to this every time for light mode, then looks like hockeystats when in dark mode. Claude Design did a more unique design but that limit was gone in 2-3 prompts?

Not quite loving the outputs from the model rn, probably needs a bit of smoothing, but we're headed in the right direction.

Couldn't recommend more!

@StatsByStas @TopDownHockey @mattdonders @AlexC_NJD @CJTDevil @yolo_pinyato At the time, I was using open-weight models because I only really needed light help on syntax. Not sure if they do similar stuff as larger commercial models. All design choices are my own as I typically prototype what I want on Canvas before implementing.

@RentoSaijo @TopDownHockey @mattdonders @AlexC_NJD @CJTDevil @yolo_pinyato Codex? I use both, but they spit out basically the same thing I thought; maybe they did train on them if both models converge to the same thing.