# US 11,443,588 B1 · Algorithmic relational odds nexus system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,443,588 B1 - Document ID: US11443588B1 - Status: Granted - Kind: B1 - Application number: 17/570,081 - Filed: 2022-01-06 - Granted: 2022-09-13 - Earliest claimed: 2022-01-06 - Assignees: Ladris Technologies, Inc. - Applicants: Leo Zlimen; Bowen Kyle - Inventors: Leo Zlimen; Bowen Kyle - Web page: https://patentgrep.com/patent/US11443588B1 ## CPC - G07F 17/323: {wherein the player is informed, e.g. advertisements, odds, instructions} - G06N 3/02: Neural networks - G06N 3/086: using evolutionary algorithms, e.g. genetic algorithms or genetic programming - G06Q 50/34: Betting or bookmaking, e.g. Internet betting - G07F 17/3288: {Betting, e.g. on live events, bookmaking} - G07F: COIN-FREED OR LIKE APPARATUS (coin sorting G07D3/00; coin testing G07D5/00; {handling coins or paper currencies apart from payment activated apparatus G07D; payment architectures, schemes or protocols G06Q20/00}) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR ## Abstract Embodiment method and associated apparatus relate to altering the expected value of a system modeled by a random process simplified to produce a binary outcome. Various embodiments modify a genetic algorithm to optimize such settings as population size, number of iterations to convergence, mutation chance, and sample space. Some embodiment ARON implementations correctly predict game outcome relative to the spread, based on transforming unrelated raw data, applying the transformed raw data to a modified genetic algorithm, generating multiple expected outcomes determined by the modified genetic algorithm as a function of the transformed raw data, and filtering the outcomes as a function of predefined metrics to produce a single end result that can be utilized effectively by an evolutionary-style algorithm. Various embodiment implementations use modified genetic algorithms with embedded neural network architecture to model and predict for a user a discrete forecast of a game-like scenario using selectively processed historical data. ## Claims ### Claim 1 An apparatus comprising: a memory storage component; a network interface; a display device; at least one processor to process at least one sports game's predictive model, and to receive and process a selection of elements for oddsmaking to predict a game's outcome, wherein the elements include one or more past game events and game-related data and historical data about a game and its historical data about its separate opposing participants; said process being: to model a game-like scenario and predict a game-like scenario outcome, the model's process comprising: a. constructing a predictive model population comprising a plurality of neural network models each configured to predict a game-like scenario outcome, wherein the game-like scenario comprises a sporting event, b. wherein each of the plurality of neural network models is configured to predict the game-like scenario outcome determined as a function of a weight value governing the influence of a neural network input on a neural network output, and c. wherein each weight value is initialized to a value determined as a function of a category of statistical data associated with the game-like scenario, and d. wherein the game-like scenario outcome is determined as a function of the measured performances of at least a first participant and a second participant in each game-like scenario, and wherein at least the first participant and the second participant in the game-like scenario are competitors; e. evolving the predictive model population of neural network models by a genetic algorithm until the game-like scenario outcomes predicted by the predictive model population of neural network models satisfy a success criterion, wherein the success criterion is determined as a function of a spread set by an oddsmaker and a Game Spread Ratio (GSR) calculated for each game-like scenario, and wherein the members of the predictive model population that satisfied the success criterion are designated as members of the parent generation which will be used to create the next generation, and f. wherein a gene in the genetic algorithm represents one neural network weight associated with data predictive of one game-like scenario participant's performance in the game-like scenario, and wherein the parent generation is filled to replace models that did not satisfy the success criterion, and wherein some of the replacement neural network models are determined as a function of the parent generation, and g. wherein some of the replacement neural network models are determined as a function of randomized data; and, h. providing access to the evolved predictive model population of neural network models to predict a future game-like scenario outcome determined as a function of historical data input, and 1. to provide a recommended betting choice from among at least the first participant and the second participant, wherein the recommended betting choice is based on predicted outcomes selectively filtered as a function of the measured performances of at least the first participant and the second participant in each game-like scenario; and wherein the processor then stores the updated game spread ratio in the memory; and wherein the apparatus' display device then displays the evolved predictive model population of neural network models to display a predicted future game outcome determined as a function of the processed historical data input, said prediction being either the winner of a game or a winner against a predetermined point spread of the game; said prediction being displayed on the apparatus' display device. ### Claim 2 A process to model a game-like scenario and predict a game-like scenario outcome, the process comprising: constructing a predictive model population comprising a plurality of neural network models each configured to predict a game-like scenario outcome, wherein the game-like scenario outcome is determined as a function of the measured performances of at least a first participant and a second participant in each game-like scenario; evolving the predictive model population of neural network models by a genetic algorithm until the game-like scenario outcomes predicted by the predictive model population of neural network models satisfy a criterion determined as a function of the measured historical performances of at least the first participant and the second participant in each game-like scenario; and, providing a user with access to the evolved predictive model population of neural network models to predict a future game-like scenario outcome determined as a function of historical data input. ### Claim 3 (depends on 2) The process of claim 2, wherein each of the plurality of neural network models is configured to predict the game-like scenario outcome determined as a function of a weight value governing the influence of a neural network input on a neural network output. ### Claim 4 (depends on 2) The process of claim 2, wherein the game-like scenario further comprises a sporting event, and wherein at least the first participant and the second participant in the game-like scenario are competitors in the sporting event, and wherein the criterion determined as a function of the measured historical performances of at least the first participant and the second participant in the game-like scenario is based on a point spread predetermined by an oddsmaker, and wherein the measured historical performances further comprise scores achieved in the sporting event. ### Claim 5 (depends on 2) The process of claim 2, wherein evolving the predictive model population by the genetic algorithm further comprises comparing the historical outcome of a game-like scenario to the outcome predicted as a function of data associated with the game-like scenario by each of the plurality of neural network models. ### Claim 6 (depends on 2) The process of claim 2, wherein evolving the predictive model population of neural network models by the genetic algorithm further comprises selectively replacing a predictive model population member failing to satisfy a success criterion, wherein the success criterion is determined as a function of the deviation of an outcome predicted by the predictive model population member from measured performances of at least the first participant and the second participant in the game-like scenario. ### Claim 7 (depends on 2) The process of claim 2, wherein evolving the predictive model population of neural network models by the genetic algorithm further comprises selectively replacing a predictive model population member failing to satisfy a success criterion with a neural network model derived from a predictive model population member satisfying the success criterion. ### Claim 8 (depends on 2) The process of claim 2, wherein evolving the predictive model population of neural network models by the genetic algorithm further comprises selectively replacing a predictive model population member failing to satisfy a success criterion with a neural network model determined as a function of randomized data selected from another predictive model population member. ### Claim 9 (depends on 2) The process of claim 2, wherein evolving the predictive model population of neural network models by a genetic algorithm further comprises identifying an error window between: a first probability that a predictive model's outcome prediction differs from a metric predetermined as a function of the measured performances of at least the first participant and the second participant; and, a second probability the metric is incorrect. ### Claim 10 (depends on 2) The process of claim 2, wherein evolving the predictive model population of neural network models by a genetic algorithm further comprises choosing a winner based on a game-like scenario predicted by the model when the performance of at least one participant in the game-like scenario predicted by the model differs from a metric predetermined as a function of the measured performances of at least the first participant and the second participant. ### Claim 11 (depends on 2) The process of claim 2, wherein predict a future game like scenario outcome determined as a function of historical data input further comprises providing the user with a recommended betting choice from among at least the first participant and the second participant, wherein the recommended betting choice is based on predicted outcomes selectively filtered as a function of the measured performances of at least the first participant and the second participant in each game-like scenario. ### Claim 12 A process to model a game-like scenario and predict a game-like scenario outcome, the process comprising: constructing a predictive model population comprising a plurality of neural network models each configured to predict a game-like scenario outcome, wherein the game-like scenario comprises a sporting event, and wherein each of the plurality of neural network models is configured to predict the game-like scenario outcome determined as a function of a weight value governing the influence of a neural network input on a neural network output, and wherein each weight value is initialized to a value determined as a function of a category of statistical data associated with the game-like scenario, and wherein the game-like scenario outcome is determined as a function of the measured performances of at least a first participant and a second participant in each game-like scenario, and wherein at least the first participant and the second participant in the game-like scenario are competitors; evolving the predictive model population of neural network models by a genetic algorithm until the game-like scenario outcomes predicted by the predictive model population of neural network models satisfy a success criterion determined as a function of a spread set by an oddsmaker and a Game Spread Ratio (GSR) calculated for each game-like scenario, wherein determining the success criterion is satisfied comprises identifying an error window between: a first probability that a predictive model's outcome prediction differs from a metric predetermined as a function of the measured performances of at least the first participant and the second participant and a second probability the metric is incorrect, and wherein the members of the predictive model population that satisfied the success criterion are designated as members of the parent generation which will be used to create the next generation, and wherein the parent generation is filled to replace models that did not satisfy the success criterion with neural network models determined as a function of randomized data; and, providing a user with access to the evolved predictive model population of neural network models to predict a future game-like scenario outcome determined as a function of historical data input, to provide the user with a recommended betting choice from among at least the first participant and the second participant, wherein the recommended betting choice is based on predicted outcomes selectively filtered as a function of the measured performances of at least the first participant and the second participant in each game-like scenario. ### Claim 13 (depends on 12) The process of claim 12, wherein constructing a predictive model population comprising a plurality of neural network models further comprises predicting a plurality of game-like scenario outcomes. ### Claim 14 (depends on 12) The process of claim 12, wherein the criterion determined as a function of a spread set by the oddsmaker and the Game Spread Ratio (GSR) is a measure of the extent to which the spread set by the oddsmaker deviates from the Game Spread Ratio (GSR). ### Claim 15 (depends on 12) The process of claim 12, wherein satisfy a criterion determined as a function of a spread set by an oddsmaker and a calculated Game Spread Ratio (GSR) further comprises an indication the predictive model is satisfactorily trained. ### Claim 16 (depends on 12) The process of claim 12, wherein evolving the predictive model population of neural network models by a genetic algorithm further comprises a gene in the genetic algorithm representing one neural network weight associated with one game-like scenario participant. ### Claim 17 A process to model a game-like scenario and predict a game-like scenario outcome, the process comprising: constructing a predictive model population comprising a plurality of neural network models each configured to predict a game-like scenario outcome, wherein the game-like scenario comprises a sporting event, and wherein each of the plurality of neural network models is configured to predict the game-like scenario outcome determined as a function of a weight value governing the influence of a neural network input on a neural network output, and wherein each weight value is initialized to a value determined as a function of a category of statistical data associated with the game-like scenario, and wherein the game-like scenario outcome is determined as a function of the measured performances of at least a first participant and a second participant in each game-like scenario, and wherein at least the first participant and the second participant in the game-like scenario are competitors; evolving the predictive model population of neural network models by a genetic algorithm until the game-like scenario outcomes predicted by the predictive model population of neural network models satisfy a success criterion, wherein the success criterion is determined as a function of a spread set by an oddsmaker and a Game Spread Ratio (GSR) calculated for each game-like scenario, and wherein the members of the predictive model population that satisfied the success criterion are designated as members of the parent generation which will be used to create the next generation, and wherein a gene in the genetic algorithm represents one neural network weight associated with data predictive of one game-like scenario participant's performance in the game-like scenario, and wherein the parent generation is filled to replace models that did not satisfy the success criterion, and wherein some of the replacement neural network models are determined as a function of the parent generation, and wherein some of the replacement neural network models are determined as a function of randomized data; and, providing a user with access to the evolved predictive model population of neural network models to predict a future game-like scenario outcome determined as a function of historical data input, to provide the user with a recommended betting choice from among at least the first participant and the second participant, wherein the recommended betting choice is based on predicted outcomes selectively filtered as a function of the measured performances of at least the first participant and the second participant in each game-like scenario. ### Claim 18 (depends on 17) The process of claim 17, wherein the recommended betting choice further comprises a choice to win the game-like scenario flat out. ### Claim 19 (depends on 17) The process of claim 17, wherein the recommended betting choice further comprises a choice against the spread set by the oddsmaker. ### Claim 20 (depends on 17) The process of claim 17, wherein the recommended betting choice further comprises a choice selected from the group consisting of first participant, second participant, and null.