Update SWARM_help.txt

MaynardMiner · MaynardMiner · Mar 8, 2020 · Feb 24, 2020 · Feb 24, 2020 · Feb 25, 2020
commit ebef0c8d99929546c71a2b9b3f4a3b7f11cb2cb3
diff --git a/help/SWARM_help.txt b/help/SWARM_help.txt
@@ -242,20 +242,26 @@ Argument          Options           Explanation
                                     the highest hashrates for each algorithm. The % values are displayed
                                     on get stats screen.
 
--Historical_Bias [0-100]            Expressed as Percentage. 
-                                    Will only affect Auto_Algo.
-                                    Will not affect niechash.
-                                    Historical_Bias bias does two things when set:
-                                     -Reduces algorithms that have no 24h returns to 100%
-                                     -Compares the deviation between 24hr estimates and 24hr
-                                      returns into a %, which then creates a EMA with each
-                                      value. (for smoothing). SWARM then applies a bonus/penality
-                                      to the algorithm based on how well it has performed
-                                      over time historically. SWARM will only apply a max % penality
-                                      or bonus of x%, where x is -Historical_Bias figure.
-                                      Deviations will be listed in stat files.
-                                      If you use, reccommended starting values are somewhere
-                                      around 25-35% historical bias.
+-Historical_Bias [penalty:bonus]    Default is 1:1. The first value is penalty modifier. Second value is
+                                    bonus modifier. Historical bias is a value that checks pools estimates
+				    vs. their actual returns, and creates a % difference between the two that
+				    is either positive (did better than predicted) or negative (did worse than 
+				    predicted). These values set the maximum penalty or bonus SWARM should give
+				    when determining profitibility. The default 1:1 sets a maximum penalty of 1%
+				    or maximum bonus of 1%. These values begin applying it to the Daily estimates
+				    and daily returns.
+				    Once SWARM begins to record more data, and generates weekly statistics, SWARM 
+				    begins to base penalty and bonus on weekly data rather than 24 hours, which is often
+				    more accurate.
+				    If you wish to only penalize a pools by a maximum of 25%, then you would use
+				      -Historical_Bias 25:1
+				    If you wish to only give a pool a bonus by a maximum of 25%, the you would use
+				      -Historical_Bias 1:25
+				    If you would like to test this stat, and are unsure where to start- 10% - 30%
+				    bonus maximums and penalties are suggested.
+				    The maximum bonus can go beyond 100%. I have seen pools return 800% that predicted
+				    on 1 coin in 24 hours.
+				    The negative maximum penalty is 100%.
 
 -Algorithm   [algorithm,algorithm,etc.] When this is specified, SWARM will not use
                                         Algorithms from pool-algos.json, and defer
@@ -677,4 +683,4 @@ Example of use;
 This will tell miner to use NVIDIA1 with Wallet1 which is BTC, and to use same wallet for CPU which is 
 BTC. It will use wallet1 when mining on nlpool, and -Nicehash_Wallet1 when mining on nicehash. 
 Both are BTC wallets. CPU miner will use 3 cpu cores.
-**************************
+**************************