diff --git a/src/essay.lua b/src/essay.lua
index 0128055..d1112f0 100644
--- a/src/essay.lua
+++ b/src/essay.lua
@@ -192,9 +192,16 @@ local function sum_levels()
 end
 local ed = ease_dollars
 function ease_dollars(mod, x)
-    ed(mod, x)
+    -- Use Talisman's to_number if available
+    local safe_mod = type(mod) == "table" and (to_number and to_number(mod) or tonumber(mod)) or mod
+    local safe_x = type(x) == "table" and (to_number and to_number(x) or tonumber(x)) or x
+    -- If conversion fails, default to 0
+    safe_mod = safe_mod or 0
+    safe_x = safe_x or 0
+    -- Call original ease_dollars with converted values
+    ed(safe_mod, safe_x)
     for i = 1, #G.jokers.cards do
-        local effects = G.jokers.cards[i]:calculate_joker({ EC_ease_dollars = mod })
+        local effects = G.jokers.cards[i]:calculate_joker({ EC_ease_dollars = safe_x })
     end
 end
 
@@ -402,10 +409,10 @@ SMODS.Joker{ --Eclipse
     end,
 
     calculate = function(self, card, context)
-        if context.cardarea == G.jokers and (sum_levels() - 12) > 0 and context.joker_main then
+        if context.cardarea == G.jokers and (sum_levels() - to_big(12)) > to_big(0) and context.joker_main then
             return {
-                message = localize{type='variable',key='a_chips',vars={(sum_levels()-12)*card.ability.extra.chip_mod}},
-                chip_mod = (sum_levels()-12)*card.ability.extra.chip_mod,
+                message = localize{type='variable',key='a_chips',vars={(sum_levels() - to_big(12))*card.ability.extra.chip_mod}},
+                chip_mod = (sum_levels() - to_big(12))*card.ability.extra.chip_mod,
                 colour = G.C.CHIPS
             }
         end