[00:00] There's been a lot of noise the lastcouple of days about Kalshi and NFL wintotals. Uh it popped up in the HammerDiscord. It's one of these stories whereif you're paying attention to the space,you kind of can't just scroll past it.

[00:13] So, I want to talk through whathappened, why it blew up the way it did,and what I think the bigger takeaway ishere. This is not as PR, not as aspokesperson, just as someone who cares

[00:24] about how these products evolve and howthey're supposed to work when somethinggoes sideways.

[00:35] >> [music]

[00:41] >> Now, before we go any further, I want toaddress the the obvious elephant in theroom here. Talshi sponsors content onthis channel, but I am not aspokesperson for them. I'm not here to

[00:53] run cover. I'm not here to spinanything. They didn't ask me to do thisvideo. Nobody requested it. This isn'tcoordinated with them in any way. Theonly reason I'm talking about it is

[01:04] because it's extremely topical rightnow. It's being discussed everywhere andit matters if you're someone who usesthese products or thinking about usingthem. And more than anything, the onething I'm not going to compromise is

[01:16] credibility. Circle off has always triedto be transparent and honest even whenit's uncomfortable, even when it'smessy, even when it's a situation thatdoesn't make anyone look great. So,

[01:27] that's how I'm going to approach thisone. Now, quick reset for anyone whohasn't used Kali before, because this isthe key to understandingwhy this whole situation played out the

[01:39] way it did. Kalshi is a predictionmarket. The simplest way to think aboutit is you're trading yes or no contractson real world outcomes. So, something

[01:50] like, will team X win over eight and ahalf games? You can buy the yes or youcan buy no. And those contracts trade atprices that reflect the market's

[02:01] probability. So if yes is trading at 60cents, that's basically the marketsaying it's around a 60% outcome. If youbuy it and it hits, it settles at $1. If

[02:12] it doesn't, it settles at zero. Andbecause it's exchangish, you can alsoget out early by selling your positionif the price moves as well. Now, here's

[02:23] the important part. Kouchi is not asports book. A sports book is bookingaction, setting lines, taking the otherside, sometimes operating under gaming

[02:35] regulations with house rules that arebuilt around voids and resettlements andand all that stuff. Kalshi operates inin a very different framework. It looks

[02:46] like betting, but it's regulated likesomething else. And that's what matterswhen things go wrong. So, here's whatactually happened. Just the facts, not

[02:56] vibes, not narratives. After the NFLregular season ended, a bunch of peoplestarted posting screenshots showing thatcertain NFL win total markets were being

[03:07] settled incorrectly. And it wasn't justone team or one weird edge case.Multiple win totals that should havebeen graded a clear yes were initially

[03:18] graded as a no, meaning a zero dollarsettlement on positions that in a normalworld are just winners.

[03:27] On top of that, some users weren't evenseeing a normal loss settlement. Theywere seeing reimbursements. Sobasically, they were getting theiroriginal stake back, but not the profit

[03:39] that they would have been owed for awinning position. So from the thecustomer's perspective, you're lookingat something that should have been paidas a winner. It's showing up as a loser,and the only thing that they're getting

[03:52] back is the amount that they put in. Aand that's obviously going to set abunch of people off. And it did. It tookoff on social media. It spread fast. itturned into a a pretty big credibility

[04:03] hit because it wasn't just, hey, onemarket is delayed. It looked like theplatform was telling winners that theydidn't win. And to be clear, I'm goingto get to where this stands now and how

[04:15] it was ultimately handled. But theinitial error and the way it surfaced tousers in real time, that was a realoperational issue and it understandablyset people off. And then you had the

[04:28] discourse part of it, which is wherethis thing went from like a, you know,messy operational issue to just fullonwildfire because Dustin Galer tweeted

[04:40] and he poured gasoline on top of thefire. And look, Galker is a voice inthis industry. He's got over 12,000followers on X. He's also been a prettyconsistent critic of Khi and prediction

[04:52] markets in general. So when he tweetssomething about this, it's going totravel fast. It's going to getscreenshotted. It's going to getreposted, quote, tweeted. People willinteract and it's going to become the

[05:04] main reference point for the story. Now,even if you don't love the framing,which honestly I I rarely love theframing of Galker stuff, the underlying

[05:14] issue is still legitimate and itdeserves attention. A settlement errorlike that, especially across multiplewin totals, it's not a small thing,

[05:26] right? It hits the core of what you'rebuying into when you use any platform,which is if I'm right, do I get paid?

[05:37] So, for me, I can separate the messengerfrom the message. You don't have to likewho brought it up to agree thatsometimes it needs to be brought up. Itneeds to be brought to the forefront.

[05:49] All right. So, this is the part thatactually matters because if you don'tunderstand what Khi is regulated as, therest of this just looks like, well, whydidn't they just press the button and

[06:00] fix it? A traditional sports book,whether it's a regulated US book oroffshore, it's operating under gamingrules and house rules. They are the

[06:12] operator. They control settlement. Whenthere's an error, they can usually go,"Okay, we graded it wrong. we're goingto avoid it or we're going to resettleit and they push the correction through.

[06:23] There's still process, there are stilldisputes, but structurally it's a lotmore flexible because it's built forthat. Prediction markets are regulated

[06:35] in in a totally different lane. Cali isunder the commodity futures/ derivativesframework which means oversight runsthrough the CFTC.

[06:47] Think of it less like a sports book andmore like an exchange for eventcontracts. You're buying and sellingcontracts that have a a definedsettlement condition. When the contract

[06:58] is settled, it's not just a a cosmeticgrade. It's an actual financial outcomethat determines who gets paid and whodoesn't. And the whole system is built

[07:09] around the idea that settlement has tobe consistent. It has to be rules-based.It has to be auditable and it has to be

[07:17] resistant to manipulation. Now, that'swhy once something is marked settled, itgets very complicated to reverse. It'snot simply, you know, we feel bad, we're

[07:30] going to change it because then youintroduce questions like, well, what'sthe formal error correction process?What precedent does this set? What doesthe audit trail look like? How do you

[07:42] ensure that you're not selectivelychanging outcomes? How do you do it in away that fits the rulebook that you'reoperating under? So, I I'm not sayingthat they're powerless. The point is

[07:53] that the mechanism for settling isdifferent. Their options can be morelimited and the process can be slowerand more formal. And and when you

[08:04] combine that with users havingsportsbook expectations, like justresettle it. I mean, that's whateveryone's saying. Just just resettle

[08:13] it. That mismatch is a big reason whythis turned into a nuclear story. Now,all of this said, this is still where Ithink Khi clearly dropped the ball. And

[08:26] I'd put it into two buckets actually.firstsettlement controls uh particularly thethe operational side because no matterwhat rulebook you operate under there

[08:37] should never be a world where a cleanobjective market like NFL win totalsgets graded wrong across every team.Like this isn't a subjective prop. It's

[08:49] not some weird stat correction thathappened after the season. It'sliterally how many games did the teamwin in the regular season. In myexperience working in the offshore

[08:59] space, auto settlement was never justone singular data feed. You'd havemultiple independent feeds that providedfinal results. And the system was set up

[09:11] so that all of them had to match for themarket to autograde. If there was anymismatch, it didn't settle and it gotkicked over to a trader where someone

[09:22] had to manually verify the outcome andthen they would grade it. And that's thebasic concept. You don't let the machineconfidently do the wrong thing at scale.

[09:34] You build in some friction on purpose.And that's what I mean when I say a amarket category like win totals, itshould have a circuit breaker likeperiod. If there's a glitch, if the data

[09:46] feed is wrong, if something is off, youhalt settlement a and you route it to tosome sort of manual review, the damagefrom being slow is always less than the

[09:57] damage from being wrong. That was just aa rule of thumb. Secondly,communication, the customer experienceside, even if the fix takes time,

[10:09] silence is unacceptable. A and fromeverything I've seen, the communicationhere was not up to par. If I'm acustomer and I'm holding a position thatshould have won and the platform is

[10:21] showing it as a loser, I don't need a aperfectly polished statement in 20minutes. But I do need something thatacknowledges reality. I need clarity on

[10:33] what's happening and what to expectnext. The message should have beenimmediate and consistent. we see theissue. Your funds are safe. You'll bemade whole. Here's what we can share

[10:46] right now. Here's when we'll update youagain. Instead, the way it came acrossto a lot of users was you gotreimbursed. End of story. And whether

[10:57] that was a customer service rep whodidn't have the right info or or aDiscord mod that was just trying tomanage chaos without authority, theeffect is the same. Nevertheless, it

[11:09] created a bad user experience because itleft people to fill in the blanksthemselves. And when people fill in theblanks around money, they naturally aregoing to assume the worst. Now, on the

[11:20] timing piece, I do want to be fair aboutone thing. Situations like this don'tget solved by the first customer servicerep who sees the ticket. They don't getsolved by whoever's answering messages

[11:33] in the Discord. This is what I wouldcall, you know, a a big decision problembecause once you've got markets that areare now incorrectly settled, you're not

[11:43] talking about flipping a few accounts.You're talking about how do we correctit inside our framework? Who has to sign

[11:53] off on it? What's the cleanest and mostcompliant way to make customers whole?And also, what does that cost? And thecost part matters. If a bunch of winnersare in incorrectly graded as losers and

[12:05] then reimbursed,making everyone whole can mean thecompany effectively has to eat a bigloss out of pocket. That's that's not asmall call. That gets escalated all the

[12:17] way up. Those conversations take time.Now, now the answer is very clear. Makethe customers whole, but it is adecision that usually needs to be madeby a select few key stakeholders. So, I

[12:29] I personally understand why it wasn't aninstant resolution, that explains thedelay. It just doesn't explain the poorcommunication because even if thedecision is still being worked through,

[12:40] you can you can tell customers, "We'reon it. We see it. You'll be made whole.Here's when we're going to update youagain." And here's the key part, becauseI don't want anyone walking away with

[12:50] the wrong takeaway here. Anyone whothinks Khi only did the right thing herebecause of PR pressure or or this wassome intentional decision to try to

[13:01] screw over customers. I mean, thesepeople are absolutely out to lunch.Imagine being in a landscape thiscompetitive right now, predictionmarkets, exchanges, you know,

[13:13] traditional sports books who are alsoentering prediction markets and thenchoosing to screw over every customerholding season win totals. It wouldnever happen. There's no world where

[13:23] that happens. Anyone acting like thatwas intentional I is either speaking inbad faith or or they're just dumb. I Ihate to put it bluntly like that, butthat's the truth. This was an

[13:36] operational failure. It was acommunication failure. Full stop. Now,now what matters for consumersis that customers were ultimately made

[13:46] whole. And that's the baselineexpectation. That's what has to happen.And this is where I want to zoom out abit because the bigger point here isn'tjust one platform had a settlement

[13:58] issue. The bigger point is how peopleshould think about betting in 2026because the landscape is way bigger thanthan just regulated book versus

[14:09] offshore. Now, this channel started asan educational betting channel. In a lotof ways, it still is. I I consider it tobe that. So, instead of just yelling

[14:18] about one incident, I'd rather use it asa real lesson in how this space works.If you're serious about betting, youwant diversification.

[14:29] You want access to liquidity. You wantaccess to different types of marketsthat might give you an edge. And youwant to understand the rules of thevenues that you're using because not

[14:41] every platform is operating the exactsame way or under the exact sameframework. But the big thing I wantpeople to understand is this. Stuff like

[14:51] this happens everywhere.It happens at regulated sports books. Ithappens offshore. It happens atexchanges. It happens in otherpeer-to-peer products. It happens in

[15:02] basically every corner of this industryat some point or another. Sometimes it'sa data feed. Sometimes it's a gradingerror. Sometimes it's a system outage.

[15:12] Sometimes it's a human mistake. None ofit is ideal. It always creates chaos inthe moment. The difference is whathappens afterwards.In a lot of places in this industry, if

[15:24] something goes wrong, you're sittingthere with zero recourse, you're arguingwith a local guy, you're dealing with anoffshore that that doesn't give a [ __ ]you're dealing with a support team thatnever responds, and you just have to eat

[15:36] it. So, so while the situation obviouslysucks and it was handled poorly inparts, the most important thing from theconsumer standpoint in my opinion ispeople were made whole. That's the

[15:49] baseline expectation. That's what has tohappen. The product category isevolving. There are going to be bumps,but at the end of the day, the consumer

[15:58] can't be the one holding the bag.Now, that said, the process still has tochange because there's always going tobe a segment of customers, especially

[16:10] the more serious ones, who evaluateplatforms, on what happens when thingsbreak. not because they're looking topile on, but because they need

[16:19] reliability to do what they do. If theexperience is confusion, silence, andbad messaging, those people are are justgoing to take their volume elsewhere.

[16:31] So, if you're a Kali, the fix isn't justpaying everyone out after the fact. It'sbuilding the safeguards and thecommunication plan so it doesn't get tothat point in the first place. So, what

[16:43] does this mean for you as someonechoosing where to bet?I think the easiest way to think aboutit is that there is this spectrum of

[16:53] risk in the industry. And it's notalways the risk people think it is. Atthe very top, you got your highest risk,your least recourse. It's always going

[17:04] to be, you know, your street and local,your your buddy's bookie, the the paperhead, whatever you want to call it. Ifthey decide they don't want to pay you

[17:14] or they decide you're too sharp, you betwith a bot, whatever it may be, theyjust disappear. There's no regulator,there's no support ticket, there's noappeals process. You're basically

[17:26] negotiating with one person and that'sit. Offshore is next. That risk map, itit's all over the place. There areoffshore books that have been around

[17:37] forever that have a reputation that theywant to protect. They'll generally doright by their customers. And there areothers where if something goes wrong,you might be shouting into the void.

[17:49] Different rules, different levels ofintegrity, different levels of recourse.You have to know who you're dealingwith. Then you have regulated sportsbooks. In general, you have more

[18:02] consumer protection there. There'soversight. There are formal complaintprocesses. There's a lot morereputational pressure. But it's also notperfect. Anyone who's been around long

[18:14] enough has seen disputes where theresolution doesn't feel fair or wherethe books lean on House rules or wherethe regulators don't necessarily come

[18:24] down the way you'd expect.I've seen many of those instances.It's safer on average, but it's also notrisk-free. And then you've got exchanges

[18:35] and prediction markets. These aredifferent. They can be great forliquidity in certain spots. They canoffer markets you can't get elsewhere.They can allow you to trade in and out,

[18:46] which is a huge feature,but they're newer products and theyoperate under different rule books. So,the risk isn't always will I get paid.

[18:57] It can be how does the process work whensomething unusual happens and howquickly does it get resolved.My personal philosophy on all of this is

[19:07] pretty simple.Risk exists everywhere.You manage it with sizing. You manage itwith diversification. You manage it by

[19:18] understanding the rules of the venueyou're using. Don't keep all of yourbankroll in one place.Don't bet amounts that assumeperfection. Don't click buttons on

[19:29] platforms that you don't understand. Ifyou approach it that way, you can takeadvantage of the best parts of differentecosystems without putting yourself in a

[19:40] spot where one issue can really ruinyour day. And just to give you mypersonal context here because I Istrongly believe transparency matters,

[19:50] I'm Canadian. I live in Ontario. I can'tpersonally use Cali dayto-day the way alot of you in the United States can. So,I'm not going to sit here and act likeI'm firing bets on it from my coach

[20:02] every Sunday like it's a normalsportsbook app for me because it isn't.That said, I do have real experiencewith the product. My betting partnerdoes market making on Koshi. So, I'm

[20:13] very familiar with how the marketsfunction, how the pricing works, whatthe actual workflow looks like whenyou're active on the platform. And infull transparency, if this happened to

[20:24] me in real time, I'd be cheesed.anyone would be. This was handled poorlyat different points and it's a bad look.But I still believe in prediction

[20:36] markets as a category. I still believein Kali in terms of being a usefulproduct for serious beters and evenrecreational bettors. I'd say that evenif they weren't a sponsor of this

[20:48] channel, that's the honest to god truth.There are real benefits in terms ofmarket access, liquidity in certainspots a and just in general the

[20:57] direction that the space is heading. Atthe same time, you're the consumer. Youdo your homework. Understand what you'reusing and the rule book that it operates

[21:08] under and make the decision that's rightfor you. If this kind of incident is adeal breaker for you, that's totallyfair. I'll wrap it up with this because

[21:19] I actually want the comments to beuseful here, not just a flame war.What's the number one thing you want tosee from any platform when somethingbreaks?

[21:29] Speed, transparency, or compensation?Pick one. Tell me why. Or if you want togo more nuts and bolts, if you wererunning a platform like this, what's the

[21:40] one safeguard you'd add so that asettlement issue can't cascade acrossmultiple mark markets? And biggerpicture, because this is where theindustry is going, do you thinkprediction markets should be treated

[21:52] more like sports books or more likefinancial exchanges? What would thatmean for dispute resolution whensomething goes wrong? I read all thecomments each week. Drop one down below.If you enjoyed the video, smash a likebutton.