Public record

@_kate_lv

Posts, receipts and the limits of the evidence.

KateBlue checkmark (X Premium subscription) @_kate_lv

Polymarket Sports9,257 followers

trading sports on @Polymarket | exploring crypto | creating content

X (opens in a new tab)

Posts bets from many betting apps. For most of the bets it posted, an account tied to it by money held the opposite side.

  • Bet on both sides
  • Several accounts
  • Paid by apps it promoted
  • Says it takes paid deals

@_kate_lv posts bets from many betting apps. Wallets attributed to apps it promoted paid its wallet, and it says openly that it takes paid deals. Money ties its Polymarket profile to a group of 21 trading accounts. These accounts often held opposite sides of the same bet, including most bets it posted. Some may belong to other people.

In fairness

It described betting both sides on two accounts in public in 2025 (‘u bet equal monies and make volume on both accs’) and listed five of its wallets in July 2026.

It says openly that it takes paid deals. Since August 2026 its main Bagel posts and a Firefly post carry X’s paid-partnership label.

Show the record5 points and their sources

What the record shows

  1. 1Two accounts in the group held opposite sides of the same bet 153 times, often bought minutes apart, about $137,000 in all. Together these pairs ended about $1,400 down.

    Post by @_kate_lv on X, 9 January 2026See it whole
    View on X (opens in a new tab)No archived copy exists

    Its biggest win, Senators vs. Avalanche, bought on 2026-01-08. Four minutes later, heooo, another profile in the group, bought the other side for $2,040 and lost it.

    2 more sources

  2. 2For 95 of the 149 bets it posted, an account in the group bought the other side, often within minutes. It mentioned a hedge in 5 of those posts and 4 replies.

    Image @_kate_lv postedSee it whole
    View on X (opens in a new tab)Archived copy (opens in a new tab)

    Its Lightning slip of 2025-12-09. The other side, the Canadiens, was bought by gracechase, another account in the group.

    2 more sources

  3. 3On 2026-07-10 its Bagel wallet bought Spain, then Belgium 40 seconds later. It posted ‘putting $1000 on Spain to Advance tonight’, then ‘just made me $370 in profit’. Both bets together lost $46.

    Image @_kate_lv postedSee it whole
    View on X (opens in a new tab)Archived copy (opens in a new tab)

    The Spain card it posted after the game.

    3 more sources

  4. 4In April 2026 it showed its main profile’s peak profit, over $10,000. Three other profiles in the group often held the other side; at that peak, the four together were about $975 down.

    Post by @_kate_lv on X, 9 April 2026See it whole
    View on X (opens in a new tab)Archived copy (opens in a new tab)

    Its post of 2026-04-09.

    4 more sources

  5. 5Its Tailgate account started with $500 from a wallet on Tailgate’s side. Minutes after a bet with that money, it wrote ‘not a payed deal’. What the $500 was for is unknown.

    Post by @_kate_lv on X, 9 May 2026See it whole
    View on X (opens in a new tab)Archived copy (opens in a new tab)

    Its reply of 2026-05-09. A week earlier it had said it talked to Tailgate’s team.

    2 more sources

Accounts tied to @_kate_lv

Payments received

Deleted posts, for example

Kate@_kate_lv · 11 Jun 2026
Deleted
Backing 🇧🇷 Brazil on @thefireflyapp for the World Cup run! Join the ⚽FIFA Prediction Festival to hunt down a share of the $50,000 prize pool. https://t.co/C4PdXJjuHO
The post on X (deleted) (opens in a new tab)Archived copy (opens in a new tab)
Kate@_kate_lv · 2 Dec 2025
Deleted
boston looks solid to me at home https://t.co/6CWsRVGnLZ
The post on X (deleted) (opens in a new tab)Archived copy (opens in a new tab)

Everything that must be said with these points

  • Who holds each key is not on the record: a common funder and a common destination show one pool of money, and a group of people who share money would leave the same receipts. The two X handles appear together in public.
  • The handle has said in public that it trades on several wallets: ‘i have my accs ready since last 2 years already’ (2025-10-15), ‘i was actively trading across 5 wallets (including my main Polymarket one)’ with five wallet cards (2026-07-20), and ‘the other acc’ (2026-08-05); its Parity card of 2026-05-29 prints the address 0x51504becd5661b144579d998ba6bf0dd735831b7. Having several accounts is not misconduct by itself, and an app account without a profile is how these apps work.
  • The strength of the link differs by tier and must be shown with it. In tier 2, three accounts (the Future-, Bitget-Wallet- and Parity-linked ones) are tied by withdrawals delivered to the katelv key, not by stakes from the declared Solana wallet; the Parity-linked one also by the card the handle posted.
  • In tier 3, six small accounts (anastaasia, andreas12, andreho, andreww, briskona, svetlikvl) bought $2,535.53 in all. Money bought per tier: katelv $71,382.84; the eleven tier-2 accounts $54,581.34; heooo $50,408.95; the eight other tier-3 accounts $25,229.16.
  • Kate0101 holds no pair with another account of the group; a profile name is free text. Its 150 USDC.e came three minutes after the profile was created from another Polymarket account, crudeoilboy, which is only the payer: 163 positions, one market in common with the group, no opposite position.
  • Six examined accounts were left out of the group (notbroken2, CryptoKesha, seyoka, infinite_temper, vuiko and one unnamed wallet).
  • These statements do not disclose any particular trade, but the method and the existence of further accounts were not kept from the audience in general; what the trading record adds is how often, and with which accounts.
  • 5 of the 95 posted positions that had an opposite position elsewhere in the group carry a hedge mention in the post itself.
  • The sentence of 2025-12-02 can be read as general advice, not as an account of when the second leg was bought.
  • In the case of 2026-08-02 the post is accurate: the second leg (heooo, MOUZ) was bought 3 hours 26 minutes after the first (the Bagel wallet, Spirit) and the stated loss (‘final loss: -$800’) is close to the record (-$843.27).
  • A word search over 63,736 posts and replies of the handle since 2025-10-01 found no further statement about hedging a trade of its own.
  • The handle described the method in public: ‘try finding like sports bets where yes and no goes for 51/49 u bet equal monies and make volume on both accs and basically won’t loose any’ (2025-10-16), after ‘i have my accs ready since last 2 years already’ under ‘we farming for poly airdrops?’ (2025-10-15); later ‘you could literally do limit orders on a multiple accs’ (2026-01-31), ‘you can always hedge from a diff wallet’ (2026-02-26), ‘i was actively trading across 5 wallets (including my main Polymarket one) while doing some paid deals’ with five wallet cards (2026-07-20), ‘so i hedged my position to cut the losses’ (2026-08-02), ‘hedged and secured $100 from the other acc’ (2026-08-05) and ‘hedge accs and sell the winning one’ (2026-08-26). The method and the existence of further accounts were therefore not kept from the audience in general; what the trading record adds is how often and with which accounts.
  • These are facts about positions; whether a posted bet had its other side shown is counted separately: in five posts the handle did mention a hedge.
  • Not all 153 pairs are the handle’s: the accounts with other owner keys may be other people, and for the 23 + 27 pairs that pair the visible profile with such accounts or involve only such accounts, two people of one group taking opposite sides by arrangement is not excluded.
  • 12 pairs win either way and 32 ended with a gain; 10 lock in a gain and 24 are open on one side, which is what hedging a real view looks like.
  • The pairs were bought from the order book: in 5 of 1,750 fills two accounts of the group were each other’s counterparty.
  • A position can sit in two pairs; the totals count each position once. The 17 flat pairs, the 33 and 61 cross-app pairs and the 17 pairs from 2026-08-04 are inside the 153 and are not added to it. The 75 % size rule is a choice made for this count; the strict screen (60 seconds, 95 %) gives 42.
  • 54 posted positions had no such pair, among them all 18 posted between 2026-03-01 and 2026-05-25.
  • The handle spoke of a hedge in the post itself for 5 of the 95 and in 4 further replies, and it described the method in public before the period began: ‘try finding like sports bets where yes and no goes for 51/49 u bet equal monies and make volume on both accs and basically won’t loose any’ (2025-10-16); later ‘you can always hedge from a diff wallet’ (2026-02-26) and ‘hedge accs and sell the winning one’ (2026-08-26).
  • butzki and gracechase never appear in the handle’s posts; all accounts have different owner keys, and for heooo, butzki and gracechase the tie is signed transfers between keys and the matched trades (and, for heooo, the handle’s own words); two or more people acting together are not excluded.
  • The legs were filled by other traders, not against each other (one shared transaction on 2025-12-03 is the exception). A pair of this kind costs about 1.6 % of the money put in (median sum of the two prices 1.016).
  • The pair rule is the rule of this count; unrelated account pairs meet such rules far less often. These pairs are the posted side of the 153, under a slightly different rule; they are not additional. No follower’s trade was examined.
  • The figures in the posts are true for the leg shown (Spain leg +$346.36, Mbappé Yes leg +$499.77).
  • In three of the four games the second leg was bought after the entry post (40 minutes, 2 hours 50 minutes and 1 hour 41 minutes later, each before the start), so those entry posts were accurate when written and it is the result posts that leave the second leg out; in Spain vs. Belgium both legs were bought before the post.
  • Buying both outcomes produces volume, which counts on the apps’ leaderboards: Bagel’s campaign payer sent this wallet 138.74 PUSD in 15 prizes between 2026-07-04 and 2026-08-01, and the handle had described the method (‘make volume on both accs and basically won’t loose any’) in a public reply on 2025-10-16. Volume prizes are an explanation the record fits, not a stated purpose.
  • The handle did disclose one hedge in this wallet (2026-06-15, ‘the only god knows why i decided to hedge’), and both wallets were among the five it listed as its own on 2026-07-20.
  • In all, the Bagel wallet bought both outcomes of 10 markets between 2026-06-13 and 2026-07-18 for $10,349.21 and ended them -$332.75; 9 posted positions had their other outcome in the same wallet (8 Bagel, 1 Firefly). They are inside the 95.
  • The profit is real for that profile, the leaderboard states it correctly, and the handle told readers to track it (‘i usually post my main bets and u can always track my polymarket page’, 2025-12-26).
  • Which leg of a pair wins cannot be chosen in advance: the katelv leg was the dearer one in 32 of 65 pairs and won 38, within what chance gives, so the record does not show that wins were steered to the visible profile.
  • The four profiles have four different owner keys; butzki and gracechase never appear in the handle’s posts; heooo’s owner key was given as a wallet by the handle once and by a second X handle 16 times. That the opposite legs belong to the same purse rests on the money links (transfers between the keys in both directions, $17,906.32 between the heooo and katelv keys and wallets; a common exchange deposit address; the run of remainder transfers of 2026-03-05) and, for heooo, on the handle’s own words; it is an inference, not a key. A group of people who pool money would leave the same record.
  • Not every loss of the three profiles is the other side of a katelv position (for heooo the pairs are 27.5 % of its positions).
  • The handle reported losses in public (‘had over 10k all time high pnl and was on a loosing break’, 2026-01-30; ‘i just lost a big ass $1800 trade’, 2026-08-02) and showed seven negative profit widgets.
  • When the copy invitation of April 2026 was posted, none of the 18 positions posted in those months was paired; the profit it advertises was made in the paired months. The handle had described the two-account volume method in a public reply on 2025-10-16.
  • The group result over the whole history (21 accounts) differs in definition from the 2025-10-25 to 2026-09-30 figures (twelve accounts); the two are kept apart. Curve points are daily values at 00:00 UTC.
  • Every leg is on Polymarket, placed through different apps; this is not hedging across rival venues.
  • An account without a profile is how these apps work. The handle posted slips from several of these apps, said it was ‘just trying out the ecosystem and different apps creating accs there’ (2026-06-18), and on 2026-07-20 listed five wallets as its own (Firefly, the main Polymarket one, Polyvaults, Bagel, Share), so the accounts’ existence was not concealed. What the chain adds is that the opposite leg sat in another app’s wallet staked by the same Solana wallet.
  • The app code says which app routed the order, not who typed it; the record does not show what the apps were aware of. The apps run volume and profit leaderboards: Firefly’s prize wallet paid the Firefly account 347.98 PUSD in 12 prize legs (300 of it on 2026-07-20 for 4th place) and nothing to the Bagel account; Bagel’s campaign payer paid the Bagel account 138.74 PUSD. Builder fees at the rates the apps publish come to about $374.24 and were earned by the apps, not paid to the handle.
  • Money for the Firefly trades came partly from Firefly: 999.82 PUSD bought with 1,000 USDC that Firefly’s hub wallet had sent to a forwarding wallet entered the Firefly account at 2026-07-03 08:09:47 UTC, the morning of the Argentina post; the declared Solana wallet funded both the Firefly account (1,109.99 in 4 transfers, 2026-06-18 to 2026-07-01) and the Bagel account (3,071.83 in 11 transfers).
  • Not every tier-2 account was staked by the declared wallet: the Tailgate account started on 500 USDC.e from a Tailgate-side key, the Future-coded one on $99.64 from an unidentified BNB-chain wallet, the Bitget-Wallet-coded one from its own owner key, and the Parity-linked account on ten payments from a fixed-amount distributor; these are tied to the handle by withdrawals delivered to the katelv key.
  • In the Knicks vs. Spurs case of 2026-06-03 the handle posted both legs, the Share loss and the Crisp win, in different threads four hours apart; neither post mentions the other. Only 2 opposite legs of posted positions were themselves posted.
  • These pairs are inside the 153 and overlap the 95. Five of the ten Bagel-Firefly pairs are World Cup markets bought before the prize of 2026-07-20.
  • What the 500 was for is not on the chain: a trading credit, a creator grant or a fee would look the same, and a starting balance supplied by the app is not necessarily what the handle meant by being paid. The handle did say it had talked to the team.
  • That the key belongs to Tailgate’s side is measured, not published by the app: the collector’s payers all trade through Tailgate; the key is the collector’s largest payee; the TailgateApp profile (0x66a06098dc496cf43803f371db917d62fa42e652) withdraws to it; it is the owner key of the Polymarket wallet 0x2d5c4b556bfb80bd634da3f4d4fa3ac794770c34. The same key paid 256 wallets, mostly whole-dollar sums, five of them within five minutes that evening (1,000.00, 1,890.00, 490.00, 2,500.00, 500.00).
  • From 2026-06-10 the declared Solana wallet staked the account with $2,379.59; the account sent $2,874.90 to the katelv key, paid Tailgate $119.78 in fees and ended -$46.09.
  • The handle says elsewhere that it takes paid deals (‘this X account has made me over $10k in two months since i started taking paid deals’, 2026-06-01).
  • X’s label can be read only for posts made from 2026-08-09, so for earlier posts the absence of disclosure can be said of the words only; the handle wrote on 2026-06-03 that one of its posts could be seen ‘as a paid partnership post’.
  • On Tailgate the handle wrote ‘didn’t get payed for this ... talked to the team earlier’ (2026-05-02) and ‘not a payed deal’ (2026-05-09); the first money in its Tailgate account came from a Tailgate-side key; the purpose is not on chain.
  • For Bagel the reply ‘not a paid promo’ dates from 2026-06-13, three months before the labelled Bagel posts; the only Bagel-side payments found are leaderboard prizes (138.74 PUSD). No payment was found for AeonArena.
  • The label tells readers that a post is a paid partnership; it does not mention that the posted Bagel positions had their other outcome on heooo.
  • The per-handle flag count here (11) is the same number the cross-handle census uses; its denominator is 160 by one count and 163 by the census.
  • What the payments were for is not on chain.
  • Firefly’s campaigns publicly offer money for posting: ‘Social rewards for sharing the event or your trading takes’ (2026-04-10), ‘$300 social rewards’ per season (2026-05-22), ‘Team Influence Rewards’ (2026-06-11) and ‘share your pick, mention us, and get a random tip’ (2026-06-18). Two of the handles repeated this in their own posts (‘$300 social rewards (for posting)’, @gusik4ever, 2026-05-29; ‘share your prediction stories → post your trades & win Firefly tips’, @_kate_lv, 2026-05-22), and the links in those posts carry numbered sharer tags (sid=), the way a share-to-earn programme counts shares.
  • The published social pool of the May season ($300) is smaller than what these wallets paid the three handles within it (300 + 200 + 150 = 650), so that pool alone does not explain the payments; an unpublished creator fee and a discretionary tip cannot be told apart. ‘Team Influence Rewards’, as @0xBumblebee itself described them, are split among users after a stage, not paid on launch day.
  • @_kate_lv told its audience on 2026-06-01 that it takes paid deals, wrote ‘everything i collab with based on prediction markets is on top of polymarket’ in a reply under its Firefly post on 2026-06-11, thanked Firefly in public for ‘the rewards’ on 2026-07-20, the same day described its Firefly trading as done ‘while doing some paid deals’, and its Firefly post of 2026-08-20 carries X’s paid-partnership label.
  • X’s paid-partnership label can be read in the public record only for posts from 2026-08-09 (raw archive captures carry the field from 03:30:15Z that day); for earlier posts the text carries no words of disclosure and whether the label was set cannot be read. The launch-week posts of June 2026 are older than that.
  • @0xBumblebee’s reply sits in a run of eight transfers the loader made in 67 minutes; a published or agreed reward for replying is as possible as a fee.
  • For @gusik4ever and @vadim_kesha1 the recipient is an address their profile withdraws to (for @gusik4ever an exchange deposit address reached through one more wallet); whose account it credits is not shown.
  • Firefly’s wallets sent 4,047.98 towards @_kate_lv: 1,500 to the owner key, 2,200.01 to the forwarding wallet and 347.98 of published prizes into its Firefly account (300 of them for 4th place, which the handle announced itself). Of the hub’s 2,200.01, 1,199.87 was swapped into the Firefly account and staked, and the rest went on to the owner key. The forwarding wallet’s 4,532.40 to the owner key is the same money path at a later point and is not added to the 4,047.98.
  • The ‘$1222 + $300’ in the handle’s post of 2026-07-20 is a trading result and a prize, not payments.
  • ‘Never routed an order through Firefly’s builder code’ says nothing: Firefly users trade through a wallet the app creates, not through their public profile.
  • 23 of the 30 whole-sum payees resolve to no profile; some may be suppliers. Published prizes (for example the 100 USDT creator awards of autumn 2025) are ordinary programme income.
  • Firefly’s wallets were not tested for payments delivered through a swap router, which sender-based ledgers miss; these lists are lower bounds.