diff --git a/tests/block_weight/block_weight.cpp b/tests/block_weight/block_weight.cpp
index 4b00fc63f..676d56006 100644
--- a/tests/block_weight/block_weight.cpp
+++ b/tests/block_weight/block_weight.cpp
@@ -138,7 +138,7 @@ static uint32_t lcg()
 
 static void test(test_t t, uint64_t blocks)
 {
-  PREFIX(10);
+  PREFIX(HF_VERSION_2021_SCALING);
 
   for (uint64_t h = 0; h < LONG_TERM_BLOCK_WEIGHT_WINDOW; ++h)
   {
@@ -177,8 +177,8 @@ static void test(test_t t, uint64_t blocks)
     }
     uint64_t ltw = bc->get_next_long_term_block_weight(w);
     cryptonote::block b;
-    b.major_version = 10;
-    b.minor_version = 10;
+    b.major_version = HF_VERSION_2021_SCALING;
+    b.minor_version = HF_VERSION_2021_SCALING;
     bc->get_db().add_block(std::make_pair(std::move(b), ""), w, ltw, bc->get_db().height(), bc->get_db().height(), {});
 
     if (!bc->update_next_cumulative_weight_limit())
diff --git a/tests/block_weight/block_weight.py b/tests/block_weight/block_weight.py
index 5ec09896d..cfd0900fb 100755
--- a/tests/block_weight/block_weight.py
+++ b/tests/block_weight/block_weight.py
@@ -18,14 +18,20 @@ ltembw = MEDIAN_THRESHOLD
 weights = [MEDIAN_THRESHOLD]*MEDIAN_WINDOW_SMALL # weights of recent blocks (B), with index -1 most recent
 lt_weights = [MEDIAN_THRESHOLD]*MEDIAN_WINDOW_BIG # long-term weights
 
+# see contrib/epee/include/misc_language.h, get_mid
+def get_mid(a, b):
+    return (a//2) + (b//2) + ((a - 2*(a//2)) + (b - 2*(b//2)))//2;
+
 # Compute the median of a list
 def get_median(vec):
-    #temp = vec
+    if len(vec) == 1:
+        return vec[0]
     temp = sorted(vec)
+    n = len(temp) // 2
     if len(temp) % 2 == 1:
-        return temp[len(temp)//2]
+        return temp[n]
     else:
-        return int((temp[len(temp)//2]+temp[len(temp)//2-1])//2)
+        return get_mid(temp[n-1], temp[n])
 
 def LCG():
   global lcg_seed
@@ -46,7 +52,7 @@ def run(t, blocks):
 
       # determine the effective weight
       stmedian = get_median(weights[-MEDIAN_WINDOW_SMALL:])
-      embw = min(max(MEDIAN_THRESHOLD,stmedian),int(MULTIPLIER_BIG*ltembw))
+      embw = min(max(ltembw,stmedian),int(MULTIPLIER_BIG*ltembw))
 
       # drop the lowest values
       weights = weights[1:]
@@ -64,7 +70,7 @@ def run(t, blocks):
       else:
         sys.exit(1)
       weights.append(max_weight)
-      lt_weights.append(min(max_weight,int(ltembw + int(ltembw * 2 / 5))))
+      lt_weights.append(min(max(max_weight, ltembw * 10 // 17),int(ltembw + int(ltembw * 7 / 10))))
 
       #print "H %u, r %u, BW %u, EMBW %u, LTBW %u, LTEMBW %u, ltmedian %u" % (block, r, max_weight, embw, lt_weights[-1], ltembw, ltmedian)
       print("H %u, BW %u, EMBW %u, LTBW %u" % (block, max_weight, embw, lt_weights[-1]))