This guy built a late-resolution bot on Polymarket and made $98,842 His algorithm almost always waits until the final seconds of 5 minute BTC Up/Down markets before building a directional position: 1. Monitors the market for most of the trading window and enters near resolution 98.2% of its purchases occur in the final minute The median first entry occurs 56 seconds before the market ends For the first four minutes, it tracks BTC’s price movements. It then starts buying its chosen outcome in small amounts 2. Buys the more likely outcome if it still considers it undervalued 89.8% of its fills are at 50c+ More than half of its purchases are at 80c+, and those trades account for 65.9% of its total volume 3. Uses the opposite outcome to adjust net exposure It bought both sides in 29.55% of markets If it holds 100 Up shares and buys 40 Down shares, part of the position becomes hedged, while 60 Up shares remain directional If it then buys another 140 Down shares, its original Up position is fully hedged, and the directional residual shifts to Down Buying the opposite outcome lets it reduce or fully reverse its directional exposure Its Polymarket account: NIULAI4 Stats: > Trades per active hour: 53.89 > Average trade size: $26.56 > Win rate: 81.42% NIULAI4’s main edge comes from late-entry probability trading: it waits for a stronger BTC signal near the end of the market, buys the undervalued outcome, and uses the opposite side to rapidly adjust its exposure if conditions change

This guy built an HFT algorithm on Polymarket with an average trade size of $34.90 Result: +$129,827 Nearly 99% of its volume comes from 5 minute BTC Up/Down markets, where it combines asynchronous pairing, MERGE, and a directional residual: 1. Accumulates both sides of the same market It bought both Up and Down in 79.6% of markets. The median gap between its first purchases of of the two outcomes is 8 seconds It uses price movements within each market to buy both sides at different times and prices 2. Frees up capital through MERGE It merges an equal number of shares from each side of the market For example: 100 Up + 80 Down 80 Up + 80 Down -> MERGE -> $80 If those matched pairs cost less than $1 each, the difference becomes a structural edge regardless of how the market resolves MERGE immediately frees up capital so it can be used again 3. Holds unpaired inventory until resolution In the example above, 20 Up shares remain after the MERGE This portion is unhedged and acts as a directional residual Its final pnl therefore depends on the edge from matched pairs, how efficiently it reuses capital, and the performance of its directional residual Its Polymarket account: flippingsharks Stats: > Trades per active hour: 73.18 > Average trade size: $34.90 > Win rate: 50.33% The bot’s main edge comes from a BUY -> pairing -> MERGE cycle: merging matched inventory frees up capital and locks in the edge on pairs bought for less than $1, while the unpaired portion stays directional until resolution

This guy built a late-resolution bot on Polymarket and made $336,662 His algorithm trades 5 and 15 minute crypto Up/Down markets using probability tracking and certainty sniping: 1. Continuously assesses how close the outcome is to being locked in At the start of each market, it records the reference price and tracks the metric used to determine the result For 5 minute BTC markets, it factors in the Chainlink TWAP, the current deviation from the reference price, the time remaining, and data from fast spot/perp feeds It then opens a small probe position while its model’s confidence is still low 2. Increases position size as the outcome becomes more certain If the probability of its chosen outcome rises, it keeps buying even at higher prices The median purchase price is 91c for the first buy and 99c for the last As the market approaches resolution and the model’s confidence rises, its exposure grows 3. Builds most of its position at 98-99c When time is running out and the final TWAP is unlikely to cross the reference price, it sharply increases its position in the outcome its model considers almost certain Its Polymarket account: BoneOhio The algorithm’s main edge comes near resolution. Once its model identifies an outcome that is almost locked in, it sharply increases its position at 98-99c

This trading bot made $363,367 on Polymarket with an average trade size of $22 and a 45% win rate I analyzed all of its trades with Claude and figured out how its algorithm works. It combines directional trading with dynamic hedging > It buys in small amounts when an outcome is cheap and its probability is low > But when its model detects a strong signal, it sharply increases its position size Its typical trading pattern looks like this: 1. BTC moves above its price at the start of the market, but prices on Polymarket have not adjusted yet 2. The algorithm already values Up above its current market price, so it starts buying and increases its position size as its confidence grows 3. If the underlying asset changes direction, it can start buying the opposite outcome as a hedge It bought both outcomes in 47.6% of markets, and those markets accounted for 70.8% of its total buy volume Its Polymarket account: bosona You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) Stats: > Trades per active hour: 109 > Average trade size: $22.31 > Win rate: 45% Its strategy is based on continuously updating fair probability estimates, scaling position size with confidence, and dynamically hedging when the underlying asset changes direction

This guy built a market-making bot and made $355,245 on Polymarket His algorithm trades 5 minute BTC Up/Down markets using complete-set accumulation and inventory rebalancing: 1. Builds complete sets for less than $1 It continuously quotes Up and Down, adjusting its order prices as its fair probability estimates and the order book change Orders on both sides do not have to fill simultaneously. It can buy Up first and Down later It uses this repricing to build complete sets for less than $1. The average cost of a matched Up + Down pair is 96.73c, edge 3.27c 2. Retains a directional residual Approximately 81% of its inventory can be matched into Up + Down pairs. The remaining 19% is kept as a directional residual in line with its model’s signal Its Polymarket account: almach You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) The main edge of this HFT algorithm comes from accumulating Up and Down at different times, while the small unpaired remainder maintains directional exposure

This guy built a trading bot on Polymarket with an average trade size of $9 Result: +$108,248 I analyzed its trades and found that its core strategy is based on dynamically accumulating both outcomes at different points throughout 5 minute BTC Up/Down markets It uses price swings within each market to buy each outcome when it becomes cheaper Here’s what that looks like in practice: > BTC declines sharply > Up falls to 42c, and the bot starts accumulating it > Then BTC rebounds > Down becomes cheaper, and the algorithm buys it at 55c Up + Down = 97c -> 3c edge These paired positions have an average edge of 2.7% But it does not always keep Up and Down perfectly balanced at a 1:1 ratio On top of its paired inventory, the algorithm often keeps a directional residual, which adds extra upside if that outcome wins Its Polymarket account: Jarvisen You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) Stats: > Trades per active hour: 680 > Average trade size: $9.11 > Win rate: 69% The main edge of this HFT algorithm comes from volatility harvesting: it alternates between accumulating Up and Down, building paired inventory with a combined cost below $1 while keeping a small directional skew toward the cheaper side

This guy built a directional/latency trading bot on Polymarket and made $249,542 His bot trades 5 and 15 minute crypto Up/Down markets using short-term momentum signals, a latency edge, and dynamic hedging: 1. Uses its own model to find mispricings The algorithm receives low-latency data on BTC, ETH, SOL, and other assets from external spot/perp feeds It tracks momentum, acceleration, and order-flow imbalance, then continuously recalculates its own fair probabilities for Up and Down If the underlying asset has already moved but the Up and Down prices still reflect the old probability estimate, it builds a position 2. Quickly reverses when the signal changes It can sell shares in the original outcome or immediately buy the opposite side as a hedge It traded both outcomes in 86.2% of markets, with a median of three full directional shifts per market Its Polymarket account: nj23adsknml3 You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) The main edge of this HFT algorithm comes from how quickly it updates its fair probabilities and how effectively it manages positions It captures a small, consistent edge and repeats the process thousands of times, steadily increasing capital

This guy built an HFT algorithm on Polymarket with an average trade size of $23.59 Result: +$246,578 The bot trades short-term crypto Up/Down markets using dynamic hedging and a directional residual: 1. First buys the outcome its model considers underpriced It continuously estimates the fair probabilities of Up and Down and compares them with prices on Polymarket The bot builds a position whenever it detects a mispricing 2. Buys the opposite side when the signal changes If the underlying asset reverses, it does not close its initial position. Instead, it starts accumulating the opposite outcome It bought both sides in 90% of markets 3. Turns part of the position into complete sets below $1 In two-sided markets, the median combined cost of Up and Down is 94.5c But this is not pure arbitrage The median inventory imbalance is 40%, so after pairing both sides, the algorithm is left with a directional residual Its Polymarket account: pspspsps5 You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) The main edge of this HFT algorithm comes from continuously shifting directional exposure as its fair probability estimates change, while purchases of the opposite outcome function either as a hedge or as part of a complete set

This guy built an HFT algorithm on Polymarket with an average trade size of $38.62 Result: +$489,494 The bot trades 5 minute BTC and ETH Up/Down markets using two-sided accumulation, complete sets, and a directional residual: 1. Starts accumulating both sides almost immediately The bot first buys one side and begins buying the opposite outcome almost immediately The median delay between its first Up and Down purchases is 7.5 seconds It then continues accumulating both sides as prices change within the market 2. Builds complete sets below $1 The main goal of the paired inventory is to achieve: VWAP(Up) + VWAP(Down) < $1 Its average combined VWAP is 97.93c, giving it an edge of 2.07c 3. Keeps a directional residual on top of the hedge The bot does not try to maintain a 1:1 balance between Up and Down In paired markets, the median ratio between the larger and smaller sides of its inventory is 1.44× If its fair probability leans more strongly toward Up or Down, the algorithm increases its inventory on that side while keeping the opposite outcome as a hedge leg Its Polymarket account: x-MoneyForWhiskas The bot’s main edge comes from combining complete sets with directional skew: the core position is hedged through Up + Down below $1, while additional inventory remains on the side its model considers underpriced

This guy built a directional/latency bot and made +$303,340 on Polymarket I analyzed its trades and found that the strategy is built around continuously re-estimating the probabilities of Up and Down in 5 minute BTC markets The bot starts trading as soon as each market opens Its model compares BTC’s movement with the opening reference price and factors in momentum, acceleration, volatility, and real-time spot/perp feeds If one outcome looks underpriced, the bot immediately starts buying it A typical sequence looks like this: > It buys Up at 55c > BTC reverses > Instead of selling its Up position, the bot starts buying Down > BTC then reverses again, and it adds more Up Buying the opposite outcome acts as a dynamic hedge and reduces net directional exposure. Some of these hedges become complete sets By the end of the market, a directional residual remains. It reflects the model’s latest probability estimate and is held until resolution Its Polymarket account: 0xAAAAA You can build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) The main edge of this HFT algorithm comes from rapidly updating its fair probability estimates and continuously shifting exposure as BTC moves, while using the opposite outcome as a dynamic hedge

This guy built an HFT bot on Polymarket and made +$129,680 The bot trades 5 minute BTC Up/Down markets using complete-set accumulation and a directional residual: 1. Continuously accumulates both sides of each market It places BUY limit orders on both Up and Down and waits for BTC movements to change the outcome prices When one side becomes cheaper, the bot gets fills on that side. As the market reprices, it starts buying the opposite outcome at prices that keep the combined cost of Up and Down below $1 The weighted average cost of its complete sets is 98.43c, giving it an edge of 1.57c 2. Keeps part of the position directional After pairing, about 21.3% of the position value remains as unbalanced inventory. It chooses which outcome to hold based on its model’s signal Its Polymarket account: nagi777 Build your own trading bot here: join.horizon.trade/retrovali… (A trial is available after signing up) The bot’s main edge comes from two layers: most of its inventory is paired into complete sets costing less than $1, while the remainder is kept as a controlled directional residual