Beta test our sick new HUD
05-09-2014
, 07:44 AM
Mod edit: Beta period complete, thread closed
-----------
I've played on-line poker for nearly 10 years and for most of that time it has been my primary source of income. I've always been driven to be on the cutting edge utilizing the technical tools available to poker players. My programming partner has a truly incredible knowledge of NoteCaddy. I was completely blown away when I saw what he could do. I felt like I was cave man rubbing sticks together while he's lighting fires with telekinesis. Our collaboration was incredibly productive as you'll soon see. This HUD is the product of thousands of hours of work by two experts and what you see in this HUD is just the start. Building this HUD took us five months of dedicated work.
This HUD was constructed with three driving principles: Innovation, accuracy and in game efficiency. We want to minimize the amount of time spent mucking around in popups looking for information while keeping your on table HUD down to a manageable size. We want the HUD data to be as accurate as possible reflection of your opponent's play verses you. Our end goal is to help you increase your win rate and hourly hand volume to maximize your hourly win rate. Having an efficient and accurate HUD is an integral part of achieving that goal.
This HUD is a new approach to HUDs. These are just the highlights, there's too much to list here. It exceeds anything I imagined a HUD could achieve.
All stats are generated by NoteCaddy. This gives us limitless flexibility designing stat parameters.
As an option, you can have a HUD for ante games which only uses data from ante games. Even when playing both ante and non-ante games vs the same opponent simultaneously you'll have HUD data for each game type.
Custom sample sizes. Instead of setting arbitrary date ranges such as hands from the last 3 months for all stats, the HUD has sample size limits for each stat. For example, rather than using every hand from the last 3 months for Btn fold vs 3B, we only use the last xxx hands where xxx hands meets our reliability threshold. High frequency spots will have stats from the last few weeks, lower frequency spots from the last few months. This allows you to see changes in your opponent's game much quicker than you could with any other HUD. Sometimes even before your opponent realizes their game has changed. The sample size for each stat is figured individually.
Timing tells. The HUD includes a graphical representation integrated into our popups showing the connection between hand strength and timing in all the major spots. Currently only available on Stars and iPoker.
Our stats, badges and range graphs are dynamic. Once the sample size for the next level reaches a level of reliability that it becomes a better reflection of your opponent's play the HUD (the on table HUD, popup stats, badges, and range graphs) automatically switches to the next level. Convergence establishes the minimum reliable sample size. Each stat, badge and range graph switch independently.
Here's what we mean by dynamic
Level One: Opponent stats, badges and range graphs reflect normalized play, using your opponent's reaction to bets/raises which fall within normal ranges and exclude all-in and effective all-in bets/raises where it would likely impact their decision. Current HUD stats reflect all actions for that spot. They lump all bet sizes, stack sizes and sometimes multiway, limped, single raised, 3B+ and raised pot play into the same stat. Our stats, badges and spark graphs reflect normalized sizing and exact situations. For example, the HUD stat fold vs flop cbet is for preflop single raised pots, heads up on the flop and eliminates undersized, oversized and all-in (including effectively all-in) betting. You will find positional, 3B, 4B and multiway cbet stats, all using similar criteria, in the HUD and popups.
Level Two: Opponent stats, badges and range graphs reflect play vs a full stacked regular (vsReg). Hands vs recreational and short stack players are excluded. Current HUDs lump together play against all player types and we all know that our play against recreational players and short stackers differs significantly. Having that play included skews the stats, reducing their accuracy. When the sample is large enough the HUD automatically switches that individual stat. Think of vsReg stats as a proxy for vsHero stats. Seeing an opponent's stats vs full stacked regulars will give you a much more accurate representation of their play vs you and reliable samples size are achieved much quicker than vsHero.
Using the previous example of fold vs flop cbet, this level will reflect the frequency an opponent folds vs a flop cbet when the cbet comes from a full stacked regular defined player in a preflop single raised pot, heads up on the flop eliminating undersized and oversized betting. In the HUD and popups you'll find positional, 3B, 4B, and multiway stats. Those stats will also switch to level two once reliability is high enough that level two provides better information.
To conserve processing power we do not use dynamic stats in those spots where it is unlikely you'll have a sample size to generate level two stats.
Level Three: Opponent stats, badges and range graphs reflects play vs you (vsHero). VsHero stats have a lot of value, but it takes a significant amount of time to search your popups for their play vs you and then you have to determine if the vsHero stat has a reliable sample size. The HUD will switch individual stats, badges and range graphs to vsHero when vsHero reaches a reliable sample size.
Badge system
Badges are HUD icons that help identify spots in an opponent's game where they are unbalanced. The badges only appear when your opponent has played exploitably over a reliable sample. While badges take some work to integrate into your game, once you do they offer a significant advantage. Badges increase your efficiency significantly by saving time looking through your popups trying to figure out if a player is exploitable in a specific spot. This is especially true when you're trying to see if your opponent is exploitable by a series of moves. Most players won't spend the time, especially in spots where many players aren't exploitable. Badges also keep the size of your on table HUD manageable.
With our HUD we take badges to another level. Playing with this HUD I found the badges were by far the most valuable feature of the package.
Currently there are 286 badges.
Bet size exploitabilities are incorporated into the badges. A modification to a badge will clearly show exploitable bet sizes. This includes vsBtn steals, vsSB steals, and vs3 bets. There are also general exploitable bet size badges for flop, turn and river.
Our badge system uses multiple stats to increase accuracy and find more exploitable spots. For example, fold BBvsBtn steal badge includes BBvsBtn preflop fold frequency , BBvsBtn preflop call frequency, and BBvsBtn fold vs flop cbet frequency. Defending by calling instead of 3B'ing adds value to the bottom of our stealing range as we'll be able to hit some flops and profitably bluff some missed flops. Often players who don't fold preflop at a clearly exploitable level will fold the flop at a very exploitable level due to a wider preflop calling range and we don't want to miss expanding our ranges verses those players. Where applicable, all our badges use multiple stats.
Badges feature two levels to reflect exploitable and highly exploitable play, allowing you to quickly determine when to extend ranges. Our badge construction is much more complex than this example, but here's a basic description: Fold BB open vs a SB steal. A level one badge (exploitable) appears with a 60-66% fold frequency and a level two badge (highly exploitable) will replace the level badge one badge when there is a 66.01%+ fold frequency. Actual badge stats includes BB fold vs SB steal frequency, BB vs SB steal calling frequency and folding BB vs SB flop cbet frequency. The exploitive level is derived from a combination of all 3 stats. If the exploitability is verses a flop cbet a second badge will activate. Instead of spending 10-15 seconds opening up a popup to find multiple stats you merely have to spend a split second looking for the badge and based on the badge (exploitable and highly exploitable) determine a range.
The data used to create a badge will switch to vsReg and vsHero when the sample size is large enough. All badges include a vsHero check. If an opponent plays non-exploitable verses you the badge will not activate.
A number of our post flop badges are combo badges revealing opponents who are exploitable verses combo moves.
Considerable effort has gone into the badge design and logic. By memorizing the fairly straight forward legend and the intuitive abbreviations you'll be able to easily identify the exploitability each badge represents. Additionally, to help identification stats are placed in your HUD by street/preflop action/situation.
Range graphs
We have a complete array of range graphs which give you a graphical representation of your opponent's play. Like the rest of our package, our spark/scatter graphs are a big leap over current spark graphs.
Our package has 375 range graphs.
Our range graphs are dynamic where applicable, switching to vsReg and vsHero when that sample is reliable.
Our range graphs feature a new graphical range representation which offers a significant improvement in interpreting how our opponents construction their ranges and the bet sizes they use. Preflop range graphs will have popup hand chart showing known range. Post flop range graphs have a popup pie chart showing known hand strengths. All the major spots have scatter graphs showing the connection between bet sizing and hand strength. These scatter graphs eliminate a huge amount of popup clutter exposing bet size exploitability.
System requirements
A significant amount of work has gone into streamlining the AJA Dynamic HUD to keep the required processor speed reasonable. This was a massive hurdle as the package currently has nearly 1900 definitions and 286 badges. The AJA Dynamic HUD requires a fast computer. We've tested the package on a number of different systems and this is what we recommend for minimum system requirements: 8GB RAM, quad core i5 2.8GHz+ CPU, SSD.
Because of the number of internal calculations and the nature of Zoom type games the Dynamic HUD often has performance issues with huge (1kk+) databases. We recommend our Premium package.
We are finishing up the beta and expect sales to start 2 weeks from this posting. Anyone is free to join the beta. You must own HM2 and NoteCaddy Premium.
Beta for the AJA Dynamic HUD.
Beta for the AJA Premium HUD.
The stake level reflects the top level that NoteCaddy will process hands. The stat information will still display at higher levels but NoteCaddy won't process hands from higher levels. If you decide to upgrade your stake level you only pay the difference between your current package and the higher stake level.
For the beta test there will be no stake limitations for the Dynamic package.
For those that don’t have a system that can handle our Dynamic HUD we offer the AJA Premium. Please note the differences below. On the surface the HUD and popups look the same for both packages. Premium lacks dynamic stats and multi-stat, integrated sizing and multi-street badges. It's still the best HUD on the market outside of our Dynamic package.
The value of an advanced HUD
Poker is game of incomplete information. The more you can fill in information gaps the better you'll be able to make decisions. The AJA Dynamic HUD fills those information gaps to a level never before seen.
However, no matter how advanced, a HUD is not going to turn a losing player into a winning player. Nothing can replace skilled poker play and mental toughness. To win in today's game at any level requires you to be a very good player with strong emotional control and no software is going to change that. Use of an advanced HUD is just another skill set for the well rounded professional. However, if properly utilized, the AJA HUD will increase your win rate. I credit 2-3bb/100 of my win rate to the edge I gain from using a NC generated HUD. Even if you can only achieve a fraction of that the AJA Dynamic HUD will pay for itself quickly. I've just started using the AJA Dynamic HUD, I'll be very interested to see what sort of gains I get out this HUD.
-----------
I've played on-line poker for nearly 10 years and for most of that time it has been my primary source of income. I've always been driven to be on the cutting edge utilizing the technical tools available to poker players. My programming partner has a truly incredible knowledge of NoteCaddy. I was completely blown away when I saw what he could do. I felt like I was cave man rubbing sticks together while he's lighting fires with telekinesis. Our collaboration was incredibly productive as you'll soon see. This HUD is the product of thousands of hours of work by two experts and what you see in this HUD is just the start. Building this HUD took us five months of dedicated work.
This HUD was constructed with three driving principles: Innovation, accuracy and in game efficiency. We want to minimize the amount of time spent mucking around in popups looking for information while keeping your on table HUD down to a manageable size. We want the HUD data to be as accurate as possible reflection of your opponent's play verses you. Our end goal is to help you increase your win rate and hourly hand volume to maximize your hourly win rate. Having an efficient and accurate HUD is an integral part of achieving that goal.
This HUD is a new approach to HUDs. These are just the highlights, there's too much to list here. It exceeds anything I imagined a HUD could achieve.
All stats are generated by NoteCaddy. This gives us limitless flexibility designing stat parameters.
As an option, you can have a HUD for ante games which only uses data from ante games. Even when playing both ante and non-ante games vs the same opponent simultaneously you'll have HUD data for each game type.
Custom sample sizes. Instead of setting arbitrary date ranges such as hands from the last 3 months for all stats, the HUD has sample size limits for each stat. For example, rather than using every hand from the last 3 months for Btn fold vs 3B, we only use the last xxx hands where xxx hands meets our reliability threshold. High frequency spots will have stats from the last few weeks, lower frequency spots from the last few months. This allows you to see changes in your opponent's game much quicker than you could with any other HUD. Sometimes even before your opponent realizes their game has changed. The sample size for each stat is figured individually.
Timing tells. The HUD includes a graphical representation integrated into our popups showing the connection between hand strength and timing in all the major spots. Currently only available on Stars and iPoker.
Our stats, badges and range graphs are dynamic. Once the sample size for the next level reaches a level of reliability that it becomes a better reflection of your opponent's play the HUD (the on table HUD, popup stats, badges, and range graphs) automatically switches to the next level. Convergence establishes the minimum reliable sample size. Each stat, badge and range graph switch independently.
Here's what we mean by dynamic
Level One: Opponent stats, badges and range graphs reflect normalized play, using your opponent's reaction to bets/raises which fall within normal ranges and exclude all-in and effective all-in bets/raises where it would likely impact their decision. Current HUD stats reflect all actions for that spot. They lump all bet sizes, stack sizes and sometimes multiway, limped, single raised, 3B+ and raised pot play into the same stat. Our stats, badges and spark graphs reflect normalized sizing and exact situations. For example, the HUD stat fold vs flop cbet is for preflop single raised pots, heads up on the flop and eliminates undersized, oversized and all-in (including effectively all-in) betting. You will find positional, 3B, 4B and multiway cbet stats, all using similar criteria, in the HUD and popups.
Level Two: Opponent stats, badges and range graphs reflect play vs a full stacked regular (vsReg). Hands vs recreational and short stack players are excluded. Current HUDs lump together play against all player types and we all know that our play against recreational players and short stackers differs significantly. Having that play included skews the stats, reducing their accuracy. When the sample is large enough the HUD automatically switches that individual stat. Think of vsReg stats as a proxy for vsHero stats. Seeing an opponent's stats vs full stacked regulars will give you a much more accurate representation of their play vs you and reliable samples size are achieved much quicker than vsHero.
Using the previous example of fold vs flop cbet, this level will reflect the frequency an opponent folds vs a flop cbet when the cbet comes from a full stacked regular defined player in a preflop single raised pot, heads up on the flop eliminating undersized and oversized betting. In the HUD and popups you'll find positional, 3B, 4B, and multiway stats. Those stats will also switch to level two once reliability is high enough that level two provides better information.
To conserve processing power we do not use dynamic stats in those spots where it is unlikely you'll have a sample size to generate level two stats.
Level Three: Opponent stats, badges and range graphs reflects play vs you (vsHero). VsHero stats have a lot of value, but it takes a significant amount of time to search your popups for their play vs you and then you have to determine if the vsHero stat has a reliable sample size. The HUD will switch individual stats, badges and range graphs to vsHero when vsHero reaches a reliable sample size.
Badge system
Badges are HUD icons that help identify spots in an opponent's game where they are unbalanced. The badges only appear when your opponent has played exploitably over a reliable sample. While badges take some work to integrate into your game, once you do they offer a significant advantage. Badges increase your efficiency significantly by saving time looking through your popups trying to figure out if a player is exploitable in a specific spot. This is especially true when you're trying to see if your opponent is exploitable by a series of moves. Most players won't spend the time, especially in spots where many players aren't exploitable. Badges also keep the size of your on table HUD manageable.
With our HUD we take badges to another level. Playing with this HUD I found the badges were by far the most valuable feature of the package.
Currently there are 286 badges.
Bet size exploitabilities are incorporated into the badges. A modification to a badge will clearly show exploitable bet sizes. This includes vsBtn steals, vsSB steals, and vs3 bets. There are also general exploitable bet size badges for flop, turn and river.
Our badge system uses multiple stats to increase accuracy and find more exploitable spots. For example, fold BBvsBtn steal badge includes BBvsBtn preflop fold frequency , BBvsBtn preflop call frequency, and BBvsBtn fold vs flop cbet frequency. Defending by calling instead of 3B'ing adds value to the bottom of our stealing range as we'll be able to hit some flops and profitably bluff some missed flops. Often players who don't fold preflop at a clearly exploitable level will fold the flop at a very exploitable level due to a wider preflop calling range and we don't want to miss expanding our ranges verses those players. Where applicable, all our badges use multiple stats.
Badges feature two levels to reflect exploitable and highly exploitable play, allowing you to quickly determine when to extend ranges. Our badge construction is much more complex than this example, but here's a basic description: Fold BB open vs a SB steal. A level one badge (exploitable) appears with a 60-66% fold frequency and a level two badge (highly exploitable) will replace the level badge one badge when there is a 66.01%+ fold frequency. Actual badge stats includes BB fold vs SB steal frequency, BB vs SB steal calling frequency and folding BB vs SB flop cbet frequency. The exploitive level is derived from a combination of all 3 stats. If the exploitability is verses a flop cbet a second badge will activate. Instead of spending 10-15 seconds opening up a popup to find multiple stats you merely have to spend a split second looking for the badge and based on the badge (exploitable and highly exploitable) determine a range.
The data used to create a badge will switch to vsReg and vsHero when the sample size is large enough. All badges include a vsHero check. If an opponent plays non-exploitable verses you the badge will not activate.
A number of our post flop badges are combo badges revealing opponents who are exploitable verses combo moves.
Considerable effort has gone into the badge design and logic. By memorizing the fairly straight forward legend and the intuitive abbreviations you'll be able to easily identify the exploitability each badge represents. Additionally, to help identification stats are placed in your HUD by street/preflop action/situation.
Range graphs
We have a complete array of range graphs which give you a graphical representation of your opponent's play. Like the rest of our package, our spark/scatter graphs are a big leap over current spark graphs.
Our package has 375 range graphs.
Our range graphs are dynamic where applicable, switching to vsReg and vsHero when that sample is reliable.
Our range graphs feature a new graphical range representation which offers a significant improvement in interpreting how our opponents construction their ranges and the bet sizes they use. Preflop range graphs will have popup hand chart showing known range. Post flop range graphs have a popup pie chart showing known hand strengths. All the major spots have scatter graphs showing the connection between bet sizing and hand strength. These scatter graphs eliminate a huge amount of popup clutter exposing bet size exploitability.
System requirements
A significant amount of work has gone into streamlining the AJA Dynamic HUD to keep the required processor speed reasonable. This was a massive hurdle as the package currently has nearly 1900 definitions and 286 badges. The AJA Dynamic HUD requires a fast computer. We've tested the package on a number of different systems and this is what we recommend for minimum system requirements: 8GB RAM, quad core i5 2.8GHz+ CPU, SSD.
Because of the number of internal calculations and the nature of Zoom type games the Dynamic HUD often has performance issues with huge (1kk+) databases. We recommend our Premium package.
We are finishing up the beta and expect sales to start 2 weeks from this posting. Anyone is free to join the beta. You must own HM2 and NoteCaddy Premium.
Beta for the AJA Dynamic HUD.
Beta for the AJA Premium HUD.
The stake level reflects the top level that NoteCaddy will process hands. The stat information will still display at higher levels but NoteCaddy won't process hands from higher levels. If you decide to upgrade your stake level you only pay the difference between your current package and the higher stake level.
For the beta test there will be no stake limitations for the Dynamic package.
For those that don’t have a system that can handle our Dynamic HUD we offer the AJA Premium. Please note the differences below. On the surface the HUD and popups look the same for both packages. Premium lacks dynamic stats and multi-stat, integrated sizing and multi-street badges. It's still the best HUD on the market outside of our Dynamic package.
The value of an advanced HUD
Poker is game of incomplete information. The more you can fill in information gaps the better you'll be able to make decisions. The AJA Dynamic HUD fills those information gaps to a level never before seen.
However, no matter how advanced, a HUD is not going to turn a losing player into a winning player. Nothing can replace skilled poker play and mental toughness. To win in today's game at any level requires you to be a very good player with strong emotional control and no software is going to change that. Use of an advanced HUD is just another skill set for the well rounded professional. However, if properly utilized, the AJA HUD will increase your win rate. I credit 2-3bb/100 of my win rate to the edge I gain from using a NC generated HUD. Even if you can only achieve a fraction of that the AJA Dynamic HUD will pay for itself quickly. I've just started using the AJA Dynamic HUD, I'll be very interested to see what sort of gains I get out this HUD.
Last edited by _dave_; 07-23-2014 at 11:11 PM.
05-26-2014
, 12:05 PM
05-26-2014
, 01:45 PM
06-14-2014
, 12:08 PM
06-30-2014
, 05:00 AM