diff --git a/Train.py b/Train.py
index e56c47eed27bd4a3b0134e4ad6e12409e520ed38..140c8960fc33b1278b624fcf01163bea2ecdf5af 100644
--- a/Train.py
+++ b/Train.py
@@ -214,8 +214,6 @@ def trainModelRl(debug, networkName, modelDir, filename, nbIter, batchSize, devF
   sentIndex = 0
 
   for epoch in range(1,nbIter+1) :
-    probaRandom = round((probas[0][0]-probas[0][2])*math.exp((-epoch+1)/probas[0][1])+probas[0][2], 2)
-    probaOracle = round((probas[1][0]-probas[1][2])*math.exp((-epoch+1)/probas[1][1])+probas[1][2], 2)
     i = 0
     totalLoss = 0.0
     while True :
@@ -231,12 +229,19 @@ def trainModelRl(debug, networkName, modelDir, filename, nbIter, batchSize, devF
       state = policy_net.extractFeatures(dicts, sentence).to(getDevice())
 
       count = 0
+      list_probas = []
+      for pb in range(len(probas)):
+        list_probas.append([round((probas[pb][0][0]-probas[pb][0][2])*math.exp((-epoch+1)/probas[pb][0][1])+probas[pb][0][2], 2),
+                           round((probas[pb][1][0]-probas[pb][1][2])*math.exp((-epoch+1)/probas[pb][1][1])+probas[pb][1][2], 2)])
 
       while True :
         missingLinks = getMissingLinks(sentence)
         transitionSet = transitionSets[sentence.state]
         fromState = sentence.state
         toState = sentence.state
+        probaRandom = list_probas[fromState][0]
+        probaOracle = list_probas[fromState][1]
+        
 
         if debug :
           sentence.printForDebug(sys.stderr)
diff --git a/main.py b/main.py
index 5b35d5ed373fade6025574086ba8ec663219d420..594e0fd606ba3f20cd61db826c939b2d54b24f24 100755
--- a/main.py
+++ b/main.py
@@ -63,6 +63,8 @@ if __name__ == "__main__" :
     help="Evolution of probability to chose action at random : (start value, decay speed, end value)")
   parser.add_argument("--probaOracle", default="0.3,2,0.0",
     help="Evolution of probability to chose action from oracle : (start value, decay speed, end value)")
+  parser.add_argument("--probaStateBack", default="0.0,20,1.0-1.0,20,0.0",
+    help="Evolution of probability to chose action in state Back with random and oracle.")
   parser.add_argument("--countBreak", default=1,
     help="Number of unaplayable transition picked before breaking the analysis.")
   args = parser.parse_args()
@@ -89,6 +91,7 @@ if __name__ == "__main__" :
     args.states = ["tagger"]
     strategy = {"TAG" : (1,0)}
     args.network = "tagger"
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))]]
   elif args.transitions == "taggerbt" :
     tmpDicts = Dicts()
     tmpDicts.readConllu(args.corpus, ["UPOS"], 0)
@@ -98,11 +101,14 @@ if __name__ == "__main__" :
     args.states = ["tagger", "backer"]
     strategy = {"TAG" : (1,1), "NOBACK" : (0,0)}
     args.network = "tagger"
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))],
+              [list(map(float, args.probaStateBack.split('-')[0].split(','))), list(map(float, args.probaStateBack.split('-')[1].split(',')))]]
   elif args.transitions == "eager" :
     transitionSets = [[Transition(elem) for elem in (["SHIFT","REDUCE","LEFT","RIGHT"]+args.ts.split(',')) if len(elem) > 0]]
     args.predictedStr = "HEAD"
     args.states = ["parser"]
     strategy = {"RIGHT" : (1,0), "SHIFT" : (1,0), "LEFT" : (0,0), "REDUCE" : (0,0)}
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))]]
   elif args.transitions == "tagparser" :
     tmpDicts = Dicts()
     tmpDicts.readConllu(args.corpus, ["UPOS"], 0)
@@ -111,6 +117,8 @@ if __name__ == "__main__" :
     args.predictedStr = "HEAD,UPOS"
     args.states = ["tagger", "parser"]
     strategy = {"RIGHT" : (1,0), "SHIFT" : (1,0), "LEFT" : (0,1), "REDUCE" : (0,1), "TAG" : (0,1)}
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))],
+              [list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))]]
   elif args.transitions == "tagparserbt" :
     tmpDicts = Dicts()
     tmpDicts.readConllu(args.corpus, ["UPOS"], 0)
@@ -119,11 +127,16 @@ if __name__ == "__main__" :
     args.predictedStr = "HEAD,UPOS"
     args.states = ["tagger", "parser", "backer"]
     strategy = {"RIGHT" : (1,2), "SHIFT" : (1,2), "LEFT" : (0,1), "REDUCE" : (0,1), "TAG" : (0,1), "NOBACK" : (0,0)}
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))],
+              [list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))],
+              [list(map(float, args.probaStateBack.split('-')[0].split(','))), list(map(float, args.probaStateBack.split('-')[1].split(',')))]]
+
   elif args.transitions == "swift" :
     transitionSets = [[Transition(elem) for elem in (["SHIFT"]+["LEFT "+str(n) for n in range(1,6)]+["RIGHT "+str(n) for n in range(1,6)]+args.ts.split(',')) if len(elem) > 0]]
     args.predictedStr = "HEAD"
     args.states = ["parser"]
     strategy = {"RIGHT" : (1,0), "SHIFT" : (1,0), "LEFT" : (0,0), "REDUCE" : (0,0)}
+    probas = [[list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))]]
   else :
     raise Exception("Unknown transition set '%s'"%args.transitions)
 
@@ -132,7 +145,6 @@ if __name__ == "__main__" :
     json.dump([args.predictedStr, [[str(t) for t in transitionSet] for transitionSet in transitionSets]], open(args.model+"/transitions.json", "w"))
     json.dump(strategy, open(args.model+"/strategy.json", "w"))
     printTS(transitionSets, sys.stderr)
-    probas = [list(map(float, args.probaRandom.split(','))), list(map(float, args.probaOracle.split(',')))]
     Train.trainMode(args.debug, args.network, args.corpus, args.type, transitionSets, strategy, args.model, int(args.iter), int(args.batchSize), args.dev, args.bootstrap, args.incr, args.reward, float(args.lr), float(args.gamma), probas, int(args.countBreak), args.predicted, args.silent)
   elif args.mode == "decode" :
     transInfos = json.load(open(args.model+"/transitions.json", "r"))