diff --git a/saiadmin-artd/src/locales/langs/en/dice/reward_config.json b/saiadmin-artd/src/locales/langs/en/dice/reward_config.json
index bc5b719..6d8d20c 100644
--- a/saiadmin-artd/src/locales/langs/en/dice/reward_config.json
+++ b/saiadmin-artd/src/locales/langs/en/dice/reward_config.json
@@ -14,6 +14,7 @@
     "colDisplayText": "Display Text",
     "colDisplayTextEn": "Display Text (EN)",
     "colRealEv": "Real Settlement",
+    "colRealReward": "Player Real Reward",
     "colTier": "Tier",
     "colRemark": "Remark",
     "placeholderTierSelect": "Tier",
@@ -53,7 +54,7 @@
     "infoNoBigwin": "No BIGWIN rows. Set tier to BIGWIN in the Reward Index tab first.",
     "btnRuleGenerate": "Generate by rules",
     "ruleGenerateTitle": "Generate reward index by rules",
-    "ruleGenerateRules": "[Generation logic (same as Create Reward Reference)]\n• 26 cells ordered by id ascending are positions 0–25; each row’s grid_number is 5–30 and unique.\n• Roll D (5–30): start at the cell whose grid_number equals D (start_index); clockwise landing = (start position + D) mod 26; counter-clockwise = start − D (if negative, +26).\n• Each reference row’s “dice points” column is the roll D; tier / real_ev / display text come from the config at the landing id.\n\n[Leopard rolls]\nFor rolls 5, 10, 15, 20, 25, 30, clockwise and counter-clockwise landing tiers must NOT be T4 or T5 (avoid leopard roll + penalty / try again).\n\n[real_ev vs tier]\nreal < −100 → T4; −100 < real < 0 → T3; 0 < real < 100 → T2; 100 < real < 500 → T1; T5 “try again” real_ev=0. Set per-tier real_ev standards below; values are written into the config and can be edited later.\n\n[Inputs]\nT1/T4/T5: fixed weighted counts. T2: minimum weighted count — clockwise and counter-clockwise must satisfy the corresponding constraints (one value per tier, applied to both directions). real_ev standards: one value per tier. Generated T1–T4 use ui_text / ui_text_en = 100 + real_ev (same EN); T5 uses “try again” / “Once again”. Remarks still distinguish break-even vs small win where applicable.",
+    "ruleGenerateRules": "[Generation logic (same as Create Reward Reference)]\n• 26 cells ordered by id ascending are positions 0–25; each row’s grid_number is 5–30 and unique.\n• Roll D (5–30): start at the cell whose grid_number equals D (start_index); clockwise landing = (start position + D) mod 26; counter-clockwise = start − D (if negative, +26).\n• Each reference row’s “dice points” column is the roll D; tier / real_ev / display text come from the config at the landing id.\n\n[Leopard rolls]\nFor rolls 5, 10, 15, 20, 25, 30, clockwise and counter-clockwise landing tiers must NOT be T4 or T5 (avoid leopard roll + penalty / once again).\n\n[Settlement amount vs tier]\nSettlement < 0 → T4; 0 < Settlement < 100 → T3; 100 < Settlement < 200 → T2; Settlement > 200 → T1; T5 “once again” settlement = 0. You can set a unified settlement standard for each tier below; generated rows write those values into the config, and details can be edited later in the table.\n\n[Inputs in this dialog]\nCount: T1/T4/T5 are fixed; T2 is minimum. Clockwise and counter-clockwise weighted counts (each roll result counts once) must each satisfy the entered values; T1, T4, and T5 are entered separately.\nSettlement standard: all cells in the same tier use the same value. On generation, T1–T4 use ui_text / ui_text_en = settlement real_ev; T5 is fixed to \"再来一次\" / \"Once again\". Remarks still distinguish break-even / small win, etc.",
     "ruleGenT1Row": "T1 (big prize)",
     "ruleGenT2Row": "T2 (small win / break-even)",
     "ruleGenT3RealEvOnly": "T3 (rake)",
@@ -63,10 +64,10 @@
     "ruleGenFixedCount": "Fixed count (CW & CCW)",
     "ruleGenRealEvStd": "real_ev standard",
     "ruleGenRealEvEditHint": "After saving, you can still edit display text, EN, real_ev and remarks per row in the table above.",
-    "ruleGenInvalidT1RealEv": "T1 real_ev must satisfy 100 < value < 500",
-    "ruleGenInvalidT2RealEv": "T2 real_ev must satisfy 0 < value < 100",
-    "ruleGenInvalidT3RealEv": "T3 real_ev must satisfy -100 < value < 0",
-    "ruleGenInvalidT4RealEv": "T4 real_ev must satisfy value < -100",
+    "ruleGenInvalidT1RealEv": "T1 settlement amount must satisfy: value > 200",
+    "ruleGenInvalidT2RealEv": "T2 settlement amount must satisfy: 100 < value < 200",
+    "ruleGenInvalidT3RealEv": "T3 settlement amount must satisfy: 0 < value < 100",
+    "ruleGenInvalidT4RealEv": "T4 settlement amount must satisfy: value < 0",
     "ruleGenInvalidT5RealEv": "T5 “try again” real_ev must be 0",
     "ruleGenT1Min": "T1 fixed count (CW & CCW)",
     "ruleGenT2Min": "T2 min (CW & CCW)",
diff --git a/saiadmin-artd/src/locales/langs/zh/dice/reward_config.json b/saiadmin-artd/src/locales/langs/zh/dice/reward_config.json
index 6ad4774..fda999f 100644
--- a/saiadmin-artd/src/locales/langs/zh/dice/reward_config.json
+++ b/saiadmin-artd/src/locales/langs/zh/dice/reward_config.json
@@ -13,7 +13,8 @@
     "colDicePoints": "色子点数",
     "colDisplayText": "显示文本",
     "colDisplayTextEn": "显示文本(英文)",
-    "colRealEv": "真实结算",
+    "colRealEv": "结算金额",
+    "colRealReward": "玩家实际中奖",
     "colTier": "所属档位",
     "colRemark": "备注",
     "placeholderTierSelect": "档位",
@@ -53,7 +54,7 @@
     "infoNoBigwin": "暂无 BIGWIN 档位配置，请先在「奖励索引」中设置 tier 为 BIGWIN",
     "btnRuleGenerate": "按规则生成",
     "ruleGenerateTitle": "按规则生成奖励索引",
-    "ruleGenerateRules": "【生成逻辑（与创建奖励对照一致）】\n• 盘面 26 格按 id 升序为位置 0～25；每条配置的 grid_number 为 5～30 且不重复。\n• 摇取点数 D（5～30）：起点为「grid_number=D」所在格位的 id（即 start_index），顺时针落点位置 = (起点位置 + D) mod 26，逆时针落点 = 起点位置 − D（若小于 0 则 +26）。\n• 对照表每条记录的「色子点数」列为摇取点数 D；档位、真实结算、显示文案取自落点格位对应 id 的配置。\n\n【豹子摇取点数】\n摇取点数为 5、10、15、20、25、30 时，其顺/逆时针落点档位不能为 T4、T5（避免对照表上出现豹子点数 + 惩罚/再来一次）。\n\n【real_ev 与 tier】\nreal < −100 → T4；−100 < real < 0 → T3；0 < real < 100 → T2；100 < real < 500 → T1；T5「再来一次」real_ev=0。下方可为各档位填写统一的 real_ev 标准，生成时写入配置；细则可稍后在表格中再改。\n\n【本弹窗输入】\n条数：T1/T4/T5「固定」；T2「不少于」——顺时针与逆时针的加权条数（每条摇取结果计一次）须分别满足所填数值；T1、T4 与 T5 分开填写。\nreal_ev 标准：同档位各格使用同一数值。生成时 T1～T4 的 ui_text / ui_text_en 均为「100+真实结算」；T5 固定为「再来一次」/「Once again」。备注仍区分完美回本/小赚等。",
+    "ruleGenerateRules": "【生成逻辑（与创建奖励对照一致）】\n• 盘面 26 格按 id 升序为位置 0～25；每条配置的 grid_number 为 5～30 且不重复。\n• 摇取点数 D（5～30）：起点为「grid_number=D」所在格位的 id（即 start_index），顺时针落点位置 = (起点位置 + D) mod 26，逆时针落点 = 起点位置 − D（若小于 0 则 +26）。\n• 对照表每条记录的「色子点数」列为摇取点数 D；档位、真实结算、显示文案取自落点格位对应 id 的配置。\n\n【豹子摇取点数】\n摇取点数为 5、10、15、20、25、30 时，其顺/逆时针落点档位不能为 T4、T5（避免对照表上出现豹子点数 + 惩罚/再来一次）。\n\n【结算金额 与 档位】\n结算金额 < 0 → T4；0 < 结算金额 < 100 → T3；100 < 结算金额 < 200 → T2；200 < 结算金额 → T1；T5「再来一次」结算金额=0。下方可为各档位填写统一的 结算金额 标准，生成时写入配置；细则可稍后在表格中再改。\n\n【本弹窗输入】\n条数：T1/T4/T5「固定」；T2「不少于」——顺时针与逆时针的加权条数（每条摇取结果计一次）须分别满足所填数值；T1、T4 与 T5 分开填写。\n结算金额 标准：同档位各格使用同一数值。生成时 T1～T4 的 显示文本 / 显示文本（英文） = 结算金额；T5 固定为「再来一次」/「Once again」。备注仍区分完美回本/小赚等。",
     "ruleGenT1Row": "T1 大奖",
     "ruleGenT2Row": "T2 小赚/回本",
     "ruleGenT3RealEvOnly": "T3 抽水",
@@ -61,13 +62,13 @@
     "ruleGenT5Row": "T5 再来一次",
     "ruleGenMinCount": "最少条数",
     "ruleGenFixedCount": "固定条数（顺/逆）",
-    "ruleGenRealEvStd": "real_ev 标准",
+    "ruleGenRealEvStd": "结算金额",
     "ruleGenRealEvEditHint": "生成并保存后，仍可在本页表格中逐条修改显示文案、英文、真实结算与备注。",
-    "ruleGenInvalidT1RealEv": "T1 的 real_ev 须满足：100 < 值 < 500",
-    "ruleGenInvalidT2RealEv": "T2 的 real_ev 须满足：0 < 值 < 100",
-    "ruleGenInvalidT3RealEv": "T3 的 real_ev 须满足：-100 < 值 < 0",
-    "ruleGenInvalidT4RealEv": "T4 的 real_ev 须满足：值 < -100",
-    "ruleGenInvalidT5RealEv": "T5「再来一次」的 real_ev 须为 0",
+    "ruleGenInvalidT1RealEv": "T1 的 结算金额 满足：200 < 值",
+    "ruleGenInvalidT2RealEv": "T2 的 结算金额 满足：100 < 值 < 200",
+    "ruleGenInvalidT3RealEv": "T3 的 结算金额 满足：0 < 值 < 100",
+    "ruleGenInvalidT4RealEv": "T4 的 结算金额 满足：值 < 0",
+    "ruleGenInvalidT5RealEv": "T5「再来一次」的 结算金额 须为 0",
     "ruleGenT1Min": "T1 固定条数（顺/逆）",
     "ruleGenT2Min": "T2 最少条数（顺/逆）",
     "ruleGenT4Max": "T4 固定条数（顺/逆）",
diff --git a/saiadmin-artd/src/views/plugin/dice/reward_config/index/index.vue b/saiadmin-artd/src/views/plugin/dice/reward_config/index/index.vue
index 5bd08a3..d923fcd 100644
--- a/saiadmin-artd/src/views/plugin/dice/reward_config/index/index.vue
+++ b/saiadmin-artd/src/views/plugin/dice/reward_config/index/index.vue
@@ -78,12 +78,18 @@
                   <template #default="{ row }">
                     <ElInputNumber
                       v-model="row.real_ev"
+                      @change="handleRealEvChange(row)"
                       controls-position="right"
                       size="small"
                       class="full-width"
                     />
                   </template>
                 </ElTableColumn>
+                <ElTableColumn :label="$t('page.configPage.colRealReward')" min-width="130" align="center">
+                  <template #default="{ row }">
+                    <span>{{ calcRealReward(row.real_ev) }}</span>
+                  </template>
+                </ElTableColumn>
                 <ElTableColumn :label="$t('page.configPage.colTier')" width="100" align="center">
                   <template #default="{ row }">
                     <ElSelect
@@ -162,6 +168,7 @@
                   <template #default="{ row }">
                     <ElInputNumber
                       v-model="row.real_ev"
+                      @change="handleRealEvChange(row)"
                       controls-position="right"
                       size="small"
                       class="full-width"
@@ -481,6 +488,24 @@
     }
   }
 
+  function calcRealReward(realEv: unknown): number {
+    const n = typeof realEv === 'number' && !Number.isNaN(realEv) ? realEv : Number(realEv)
+    if (Number.isNaN(n)) {
+      return -100
+    }
+    return n - 100
+  }
+
+  function handleRealEvChange(row: IndexRow) {
+    const n =
+      typeof row.real_ev === 'number' && !Number.isNaN(row.real_ev)
+        ? row.real_ev
+        : Number(row.real_ev)
+    const text = Number.isNaN(n) ? '' : String(n)
+    row.ui_text = text
+    row.ui_text_en = text
+  }
+
   async function handleCreateRewardReference() {
     try {
       await ElMessageBox.confirm(
diff --git a/saiadmin-artd/src/views/plugin/dice/reward_config/index/modules/edit-dialog.vue b/saiadmin-artd/src/views/plugin/dice/reward_config/index/modules/edit-dialog.vue
index 69350c5..3a95809 100644
--- a/saiadmin-artd/src/views/plugin/dice/reward_config/index/modules/edit-dialog.vue
+++ b/saiadmin-artd/src/views/plugin/dice/reward_config/index/modules/edit-dialog.vue
@@ -22,7 +22,11 @@
         <el-input v-model="formData.ui_text_en" :placeholder="$t('page.form.placeholderUiTextEn')" />
       </el-form-item>
       <el-form-item :label="$t('page.form.labelRealEv')" prop="real_ev">
-        <el-input-number v-model="formData.real_ev" :placeholder="$t('page.form.placeholderRealEv')" />
+        <el-input-number
+          v-model="formData.real_ev"
+          :placeholder="$t('page.form.placeholderRealEv')"
+          @change="handleRealEvChange"
+        />
       </el-form-item>
       <el-form-item :label="$t('page.form.labelTier')" prop="tier">
         <el-select
@@ -237,6 +241,16 @@
       console.log('表单验证失败:', error)
     }
   }
+
+  const handleRealEvChange = () => {
+    const n =
+      typeof formData.real_ev === 'number' && !Number.isNaN(formData.real_ev)
+        ? formData.real_ev
+        : Number(formData.real_ev)
+    const text = Number.isNaN(n) ? '' : String(n)
+    formData.ui_text = text
+    formData.ui_text_en = text
+  }
 </script>
 
 <style lang="scss" scoped>
diff --git a/saiadmin-artd/src/views/plugin/dice/reward_config/utils/generateIndexByRules.ts b/saiadmin-artd/src/views/plugin/dice/reward_config/utils/generateIndexByRules.ts
index 895da33..a186bab 100644
--- a/saiadmin-artd/src/views/plugin/dice/reward_config/utils/generateIndexByRules.ts
+++ b/saiadmin-artd/src/views/plugin/dice/reward_config/utils/generateIndexByRules.ts
@@ -53,12 +53,12 @@ export interface TierRealEvStandards {
   T5: number
 }
 
-/** 默认标准（与常见业务约定一致：100&lt;T1&lt;500，等） */
+/** 默认标准（与规则弹窗说明一致） */
 export const DEFAULT_TIER_REAL_EV_STANDARDS: TierRealEvStandards = {
-  T1: 400,
-  T2: 50,
-  T3: -80,
-  T4: -140,
+  T1: 300,
+  T2: 150,
+  T3: 50,
+  T4: -40,
   T5: 0
 }
 
@@ -66,16 +66,16 @@ export const DEFAULT_TIER_REAL_EV_STANDARDS: TierRealEvStandards = {
  * 校验档位与 real_ev 区间是否一致；通过返回 null，否则返回 i18n 键名（不含 page.configPage. 前缀）
  */
 export function validateTierRealEvStandards(s: TierRealEvStandards): string | null {
-  if (!Number.isFinite(s.T1) || !(s.T1 > 100 && s.T1 < 500)) {
+  if (!Number.isFinite(s.T1) || !(s.T1 > 200)) {
     return 'ruleGenInvalidT1RealEv'
   }
-  if (!Number.isFinite(s.T2) || !(s.T2 > 0 && s.T2 < 100)) {
+  if (!Number.isFinite(s.T2) || !(s.T2 > 100 && s.T2 < 200)) {
     return 'ruleGenInvalidT2RealEv'
   }
-  if (!Number.isFinite(s.T3) || !(-100 < s.T3 && s.T3 < 0)) {
+  if (!Number.isFinite(s.T3) || !(s.T3 > 0 && s.T3 < 100)) {
     return 'ruleGenInvalidT3RealEv'
   }
-  if (!Number.isFinite(s.T4) || !(s.T4 < -100)) {
+  if (!Number.isFinite(s.T4) || !(s.T4 < 0)) {
     return 'ruleGenInvalidT4RealEv'
   }
   if (!Number.isFinite(s.T5) || s.T5 !== 0) {
@@ -271,10 +271,21 @@ export function generateTiers(input: GenerateTierInput): GenerateTierResult {
   return { ok: false, message: '在当前盘面与约束下未找到可行解，请调整 T1/T4/T5 固定条数或放宽 T2 下限后重试' }
 }
 
+function uiTextByTierWhenStandards(
+  tier: IndexTier,
+  realEv: number
+): { ui_text: string; ui_text_en: string } {
+  if (tier === 'T5') {
+    return { ui_text: '再来一次', ui_text_en: 'Once again' }
+  }
+  const value = String(realEv)
+  return { ui_text: value, ui_text_en: value }
+}
+
 /** 展示文案：直接使用真实结算值（中英文相同） */
 function uiTextFromRealEv(realEv: number): { ui_text: string; ui_text_en: string } {
-  const s = String(realEv)
-  return { ui_text: s, ui_text_en: s }
+  const value = String(realEv)
+  return { ui_text: value, ui_text_en: value }
 }
 
 /**
@@ -315,31 +326,31 @@ export function buildRowsFromTiers(
     if (standards !== undefined) {
       if (tier === 'T1') {
         real_ev = standards.T1
-        const f = uiTextFromRealEv(real_ev)
+        const f = uiTextByTierWhenStandards(tier, real_ev)
         ui_text = f.ui_text
         ui_text_en = f.ui_text_en
         remark = '大奖格'
       } else if (tier === 'T2') {
         real_ev = standards.T2
-        const f = uiTextFromRealEv(real_ev)
+        const f = uiTextByTierWhenStandards(tier, real_ev)
         ui_text = f.ui_text
         ui_text_en = f.ui_text_en
         remark = standards.T2 <= 1 ? '完美回本' : '小赚'
       } else if (tier === 'T3') {
         real_ev = standards.T3
-        const f = uiTextFromRealEv(real_ev)
+        const f = uiTextByTierWhenStandards(tier, real_ev)
         ui_text = f.ui_text
         ui_text_en = f.ui_text_en
         remark = '抽水'
       } else if (tier === 'T4') {
         real_ev = standards.T4
-        const f = uiTextFromRealEv(real_ev)
+        const f = uiTextByTierWhenStandards(tier, real_ev)
         ui_text = f.ui_text
         ui_text_en = f.ui_text_en
         remark = '惩罚'
       } else {
         real_ev = standards.T5
-        const f = uiTextFromRealEv(real_ev)
+        const f = uiTextByTierWhenStandards(tier, real_ev)
         ui_text = f.ui_text
         ui_text_en = f.ui_text_en
         remark = '前端需要在播放一次动画(特殊)'